H0.11 top features

Top feature 0 in H0.11: (feature 12420

TOP ACTIVATIONS
MAX = 2.212

m
Tokenm
Feature activation+0.000
gonna
Token gonna
Feature activation+0.000
call
Token call
Feature activation+0.000
Ted
Token Ted
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.173
Posted
TokenPosted
Feature activation+2.212
in
Token in
Feature activation+1.065
Ċ
TokenĊ
Feature activation+0.300
Ċ
TokenĊ
Feature activation+0.062
Dear
TokenDear
Feature activation+0.920
friends
Token friends
Feature activation+0.678
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
going
Token going
Feature activation+0.000
forward
Token forward
Feature activation+0.000
."
Token."
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
Dead
TokenDead
Feature activation+2.027
hungry
Token hungry
Feature activation+1.348
:
Token:
Feature activation+0.712
Mother
Token Mother
Feature activation+1.295
-
Token-
Feature activation+0.548
to
Tokento
Feature activation+0.668
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
n
Tokenn
Feature activation+0.000
ett
Tokenett
Feature activation+0.000
.
Token.
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.086
US
TokenUS
Feature activation+1.914
exceptional
Token exceptional
Feature activation+1.672
ism
Tokenism
Feature activation+1.094
rhetoric
Token rhetoric
Feature activation+1.479
poses
Token poses
Feature activation+0.912
extreme
Token extreme
Feature activation+0.709
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
with
Token with
Feature activation+0.000
less
Token less
Feature activation+0.000
power
Token power
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.002
Sh
TokenSh
Feature activation+1.886
ining
Tokenining
Feature activation+1.789
Reson
Token Reson
Feature activation+1.401
ance
Tokenance
Feature activation+1.242
Re
Token Re
Feature activation+1.332
:
Token:
Feature activation+0.653
band
Token band
Feature activation+0.000
ing
Tokening
Feature activation+0.000
together
Token together
Feature activation+0.000
,
Token,
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
After
TokenAfter
Feature activation+1.861
talking
Token talking
Feature activation+1.501
about
Token about
Feature activation+1.069
how
Token how
Feature activation+0.830
unlikely
Token unlikely
Feature activation+0.853
it
Token it
Feature activation+0.618
sc
Tokensc
Feature activation+0.000
ast
Tokenast
Feature activation+0.000
unit
Token unit
Feature activation+0.000
:
Token:
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.158
Wars
TokenWars
Feature activation+1.824
were
Token were
Feature activation+1.328
fought
Token fought
Feature activation+1.175
to
Token to
Feature activation+0.423
impose
Token impose
Feature activation+0.875
opium
Token opium
Feature activation+0.764
to
Token to
Feature activation+0.000
this
Token this
Feature activation+0.000
report
Token report
Feature activation+0.000
.
Token.
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
No
TokenNo
Feature activation+1.823
self
Token self
Feature activation+1.319
-
Token-
Feature activation+0.752
respect
Tokenrespect
Feature activation+1.075
ing
Tokening
Feature activation+0.935
olig
Token olig
Feature activation+0.640
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
and
Token and
Feature activation+0.000
community
Token community
Feature activation+0.000
groups
Token groups
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
At
TokenAt
Feature activation+1.811
President
Token President
Feature activation+1.482
Donald
Token Donald
Feature activation+1.133
Trump
Token Trump
Feature activation+0.664
's
Token's
Feature activation+0.638
Tuesday
Token Tuesday
Feature activation+0.862
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
.
Token.
Feature activation+0.000
m
Tokenm
Feature activation+0.000
.
Token.
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
As
TokenAs
Feature activation+1.791
good
Token good
Feature activation+1.447
as
Token as
Feature activation+1.156
Google
Token Google
Feature activation+1.002
Maps
Token Maps
Feature activation+0.706
is
Token is
Feature activation+0.568
with
Token with
Feature activation+0.000
less
Token less
Feature activation+0.000
power
Token power
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.002
Sh
TokenSh
Feature activation+1.886
ining
Tokenining
Feature activation+1.789
Reson
Token Reson
Feature activation+1.401
ance
Tokenance
Feature activation+1.242
Re
Token Re
Feature activation+1.332
:
Token:
Feature activation+0.653
f
Tokenf
Feature activation+0.580
izers
Tokenizers
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ŀ
TokenĿ
Feature activation+0.000
âĢĵ
Token âĢĵ
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
Due
TokenDue
Feature activation+1.782
to
Token to
Feature activation+0.854
boring
Token boring
Feature activation+1.211
circumstances
Token circumstances
Feature activation+0.981
beyond
Token beyond
Feature activation+0.628
my
Token my
Feature activation+0.767
com
Tokencom
Feature activation+0.000
has
Token has
Feature activation+0.000
been
Token been
Feature activation+0.000
occupied
Token occupied
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
Pre
TokenPre
Feature activation+1.778
face
Tokenface
Feature activation+1.256
The
Token The
Feature activation+0.887
essence
Token essence
Feature activation+1.026
of
Token of
Feature activation+0.181
the
Token the
Feature activation+0.111
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
the
Token the
Feature activation+0.000
Lab
Token Lab
Feature activation+0.000
.
Token.
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
Half
TokenHalf
Feature activation+1.775
of
Token of
Feature activation+0.899
the
Token the
Feature activation+0.775
residents
Token residents
Feature activation+0.975
of
Token of
Feature activation+0.482
Illinois
Token Illinois
Feature activation+0.755
Like
Token Like
Feature activation+0.000
Loading
Token Loading
Feature activation+0.000
...
Token...
Feature activation+0.000
Related
Token Related
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
5
Token5
Feature activation+1.763
.
Token.
Feature activation+0.449
0
Token0
Feature activation+1.309
âĺħ
Token âĺħ
Feature activation+0.891
âĺħâĺħ
Tokenâĺħâĺħ
Feature activation+0.527
âĺħâĺħ
Tokenâĺħâĺħ
Feature activation+0.472
al
Tokenal
Feature activation+0.000
ane
Tokenane
Feature activation+0.000
z
Tokenz
Feature activation+0.000
@
Token@
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.104
H
TokenH
Feature activation+1.742
ollywood
Tokenollywood
Feature activation+1.411
's
Token's
Feature activation+1.023
highest
Token highest
Feature activation+1.018
profile
Token profile
Feature activation+0.734
feminist
Token feminist
Feature activation+0.580
W
TokenW
Feature activation+0.000
is
Tokenis
Feature activation+0.000
eman
Tokeneman
Feature activation+0.000
AP
TokenAP
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.138
Air
TokenAir
Feature activation+1.733
guns
Token guns
Feature activation+1.376
used
Token used
Feature activation+1.235
for
Token for
Feature activation+0.697
marine
Token marine
Feature activation+0.881
oil
Token oil
Feature activation+0.563
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
First
TokenFirst
Feature activation+0.000
posted
Token posted
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.004
It
TokenIt
Feature activation+1.733
âĢ
TokenâĢ
Feature activation+1.035
Ļ
TokenĻ
Feature activation+1.049
s
Tokens
Feature activation+0.714
time
Token time
Feature activation+0.729
for
Token for
Feature activation+0.406
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
buyers
Token buyers
Feature activation+0.000
guides
Token guides
Feature activation+0.000
.
Token.
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
More
TokenMore
Feature activation+1.731
than
Token than
Feature activation+1.309
138
Token 138
Feature activation+1.091
million
Token million
Feature activation+0.602
people
Token people
Feature activation+0.524
,
Token,
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
the
Token the
Feature activation+0.000
health
Token health
Feature activation+0.000
department
Token department
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
Hor
TokenHor
Feature activation+1.717
ace
Tokenace
Feature activation+1.525
Augustus
Token Augustus
Feature activation+1.071
Curtis
Token Curtis
Feature activation+1.129
VC
Token VC
Feature activation+0.904
(
Token (
Feature activation+0.424
the
Token the
Feature activation+0.000
o
Token o
Feature activation+0.000
lf
Tokenlf
Feature activation+0.000
actory
Tokenactory
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
Both
TokenBoth
Feature activation+1.706
of
Token of
Feature activation+0.934
these
Token these
Feature activation+1.065
are
Token are
Feature activation+0.679
fix
Token fix
Feature activation+0.726
able
Tokenable
Feature activation+0.543

Top DFA by src position
MAX = 2.942

<|endoftext|>
Token<|endoftext|>
Feature activation+0.362
Top resid features:
m
Tokenm
Feature activation+0.050
Top resid features:
gonna
Token gonna
Feature activation+0.130
Top resid features:
call
Token call
Feature activation+0.066
Top resid features:
Ted
Token Ted
Feature activation+0.125
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+2.942
Top resid features:
Posted
TokenPosted
Feature activation+0.712
Top resid features:
in
Token in
Feature activation+0.000
Top resid features:
Ċ
TokenĊ
Feature activation+0.000
Top resid features:
Ċ
TokenĊ
Feature activation+0.000
Top resid features:
Dear
TokenDear
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.418
Top resid features:
going
Token going
Feature activation+0.063
Top resid features:
forward
Token forward
Feature activation+0.085
Top resid features:
."
Token."
Feature activation+0.219
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+2.565
Top resid features:
Dead
TokenDead
Feature activation+0.851
Top resid features:
hungry
Token hungry
Feature activation+0.000
Top resid features:
:
Token:
Feature activation+0.000
Top resid features:
Mother
Token Mother
Feature activation+0.000
Top resid features:
-
Token-
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.482
Top resid features:
n
Tokenn
Feature activation-0.007
Top resid features:
ett
Tokenett
Feature activation+0.033
Top resid features:
.
Token.
Feature activation+0.155
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+2.592
Top resid features:
US
TokenUS
Feature activation+0.835
Top resid features:
exceptional
Token exceptional
Feature activation+0.000
Top resid features:
ism
Tokenism
Feature activation+0.000
Top resid features:
rhetoric
Token rhetoric
Feature activation+0.000
Top resid features:
poses
Token poses
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.430
Top resid features:
with
Token with
Feature activation+0.129
Top resid features:
less
Token less
Feature activation+0.103
Top resid features:
power
Token power
Feature activation+0.056
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+2.658
Top resid features:
Sh
TokenSh
Feature activation+0.685
Top resid features:
ining
Tokenining
Feature activation+0.000
Top resid features:
Reson
Token Reson
Feature activation+0.000
Top resid features:
ance
Tokenance
Feature activation+0.000
Top resid features:
Re
Token Re
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.421
Top resid features:
band
Token band
Feature activation-0.012
Top resid features:
ing
Tokening
Feature activation+0.024
Top resid features:
together
Token together
Feature activation+0.051
Top resid features:
,
Token,
Feature activation+0.097
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+2.539
Top resid features:
After
TokenAfter
Feature activation+0.916
Top resid features:
talking
Token talking
Feature activation+0.000
Top resid features:
about
Token about
Feature activation+0.000
Top resid features:
how
Token how
Feature activation+0.000
Top resid features:
unlikely
Token unlikely
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.417
Top resid features:
sc
Tokensc
Feature activation+0.009
Top resid features:
ast
Tokenast
Feature activation-0.012
Top resid features:
unit
Token unit
Feature activation+0.053
Top resid features:
:
Token:
Feature activation+0.114
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+2.833
Top resid features:
Wars
TokenWars
Feature activation+0.586
Top resid features:
were
Token were
Feature activation+0.000
Top resid features:
fought
Token fought
Feature activation+0.000
Top resid features:
to
Token to
Feature activation+0.000
Top resid features:
impose
Token impose
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.414
Top resid features:
to
Token to
Feature activation+0.069
Top resid features:
this
Token this
Feature activation+0.090
Top resid features:
report
Token report
Feature activation+0.064
Top resid features:
.
Token.
Feature activation+0.131
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+2.234
Top resid features:
No
TokenNo
Feature activation+0.997
Top resid features:
self
Token self
Feature activation+0.000
Top resid features:
-
Token-
Feature activation+0.000
Top resid features:
respect
Tokenrespect
Feature activation+0.000
Top resid features:
ing
Tokening
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.438
Top resid features:
and
Token and
Feature activation+0.086
Top resid features:
community
Token community
Feature activation+0.048
Top resid features:
groups
Token groups
Feature activation+0.050
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+2.450
Top resid features:
At
TokenAt
Feature activation+0.914
Top resid features:
President
Token President
Feature activation+0.000
Top resid features:
Donald
Token Donald
Feature activation+0.000
Top resid features:
Trump
Token Trump
Feature activation+0.000
Top resid features:
's
Token's
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.433
Top resid features:
.
Token.
Feature activation+0.106
Top resid features:
m
Tokenm
Feature activation+0.054
Top resid features:
.
Token.
Feature activation+0.144
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+2.403
Top resid features:
As
TokenAs
Feature activation+0.826
Top resid features:
good
Token good
Feature activation+0.000
Top resid features:
as
Token as
Feature activation+0.000
Top resid features:
Google
Token Google
Feature activation+0.000
Top resid features:
Maps
Token Maps
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.393
Top resid features:
with
Token with
Feature activation+0.133
Top resid features:
less
Token less
Feature activation+0.133
Top resid features:
power
Token power
Feature activation+0.040
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+2.233
Top resid features:
Sh
TokenSh
Feature activation+0.275
Top resid features:
ining
Tokenining
Feature activation+0.756
Top resid features:
Reson
Token Reson
Feature activation+0.000
Top resid features:
ance
Tokenance
Feature activation+0.000
Top resid features:
Re
Token Re
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.343
Top resid features:
izers
Tokenizers
Feature activation+0.004
Top resid features:
âĢ
TokenâĢ
Feature activation+0.100
Top resid features:
Ŀ
TokenĿ
Feature activation+0.022
Top resid features:
âĢĵ
Token âĢĵ
Feature activation+0.262
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+2.423
Top resid features:
Due
TokenDue
Feature activation+0.802
Top resid features:
to
Token to
Feature activation+0.000
Top resid features:
boring
Token boring
Feature activation+0.000
Top resid features:
circumstances
Token circumstances
Feature activation+0.000
Top resid features:
beyond
Token beyond
Feature activation+0.000
Top resid features:
.
Token.
Feature activation+0.122
Top resid features:
com
Tokencom
Feature activation+0.077
Top resid features:
has
Token has
Feature activation+0.080
Top resid features:
been
Token been
Feature activation+0.085
Top resid features:
occupied
Token occupied
Feature activation+0.073
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+2.166
Top resid features:
Pre
TokenPre
Feature activation+0.936
Top resid features:
face
Tokenface
Feature activation+0.000
Top resid features:
The
Token The
Feature activation+0.000
Top resid features:
essence
Token essence
Feature activation+0.000
Top resid features:
of
Token of
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.394
Top resid features:
the
Token the
Feature activation+0.035
Top resid features:
Lab
Token Lab
Feature activation+0.015
Top resid features:
.
Token.
Feature activation+0.199
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+2.449
Top resid features:
Half
TokenHalf
Feature activation+0.859
Top resid features:
of
Token of
Feature activation+0.000
Top resid features:
the
Token the
Feature activation+0.000
Top resid features:
residents
Token residents
Feature activation+0.000
Top resid features:
of
Token of
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.275
Top resid features:
Like
Token Like
Feature activation+0.055
Top resid features:
Loading
Token Loading
Feature activation+0.158
Top resid features:
...
Token...
Feature activation+0.116
Top resid features:
Related
Token Related
Feature activation+0.117
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+2.366
Top resid features:
5
Token5
Feature activation+0.852
Top resid features:
.
Token.
Feature activation+0.000
Top resid features:
0
Token0
Feature activation+0.000
Top resid features:
âĺħ
Token âĺħ
Feature activation+0.000
Top resid features:
âĺħâĺħ
Tokenâĺħâĺħ
Feature activation+0.000
Top resid features:
te
Tokente
Feature activation+0.034
Top resid features:
al
Tokenal
Feature activation+0.057
Top resid features:
ane
Tokenane
Feature activation+0.029
Top resid features:
z
Tokenz
Feature activation+0.024
Top resid features:
@
Token@
Feature activation+0.118
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+2.569
Top resid features:
H
TokenH
Feature activation+0.758
Top resid features:
ollywood
Tokenollywood
Feature activation+0.000
Top resid features:
's
Token's
Feature activation+0.000
Top resid features:
highest
Token highest
Feature activation+0.000
Top resid features:
profile
Token profile
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.397
Top resid features:
W
TokenW
Feature activation+0.063
Top resid features:
is
Tokenis
Feature activation+0.035
Top resid features:
eman
Tokeneman
Feature activation+0.024
Top resid features:
AP
TokenAP
Feature activation+0.076
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+2.684
Top resid features:
Air
TokenAir
Feature activation+0.630
Top resid features:
guns
Token guns
Feature activation+0.000
Top resid features:
used
Token used
Feature activation+0.000
Top resid features:
for
Token for
Feature activation+0.000
Top resid features:
marine
Token marine
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.359
Top resid features:
Ċ
TokenĊ
Feature activation+0.116
Top resid features:
First
TokenFirst
Feature activation+0.169
Top resid features:
posted
Token posted
Feature activation+0.073
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+2.385
Top resid features:
It
TokenIt
Feature activation+0.805
Top resid features:
âĢ
TokenâĢ
Feature activation+0.000
Top resid features:
Ļ
TokenĻ
Feature activation+0.000
Top resid features:
s
Tokens
Feature activation+0.000
Top resid features:
time
Token time
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.341
Top resid features:
buyers
Token buyers
Feature activation-0.004
Top resid features:
guides
Token guides
Feature activation+0.021
Top resid features:
.
Token.
Feature activation+0.188
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+2.402
Top resid features:
More
TokenMore
Feature activation+0.958
Top resid features:
than
Token than
Feature activation+0.000
Top resid features:
138
Token 138
Feature activation+0.000
Top resid features:
million
Token million
Feature activation+0.000
Top resid features:
people
Token people
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.407
Top resid features:
the
Token the
Feature activation+0.094
Top resid features:
health
Token health
Feature activation+0.088
Top resid features:
department
Token department
Feature activation+0.084
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+2.538
Top resid features:
Hor
TokenHor
Feature activation+0.680
Top resid features:
ace
Tokenace
Feature activation+0.000
Top resid features:
Augustus
Token Augustus
Feature activation+0.000
Top resid features:
Curtis
Token Curtis
Feature activation+0.000
Top resid features:
VC
Token VC
Feature activation+0.000
Top resid features:
influence
Token influence
Feature activation-0.004
Top resid features:
the
Token the
Feature activation+0.055
Top resid features:
o
Token o
Feature activation+0.027
Top resid features:
lf
Tokenlf
Feature activation+0.070
Top resid features:
actory
Tokenactory
Feature activation+0.061
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+2.324
Top resid features:
Both
TokenBoth
Feature activation+0.929
Top resid features:
of
Token of
Feature activation+0.000
Top resid features:
these
Token these
Feature activation+0.000
Top resid features:
are
Token are
Feature activation+0.000
Top resid features:
fix
Token fix
Feature activation+0.000
Top resid features:

Decoder Weights Distribution

Head 0: 0.07

Head 1: 0.09

Head 2: 0.08

Head 3: 0.09

Head 4: 0.03

Head 5: 0.03

Head 6: 0.09

Head 7: 0.08

Head 8: 0.06

Head 9: 0.17

Head 10: 0.06

Head 11: 0.16

Positive logits

terness3.22

3.04

atform2.97

VIDIA2.72

NetMessage2.66

��2.63

terday2.56

ccording2.55

cipled2.48

NESS2.45

tesy2.45

oshop2.41

lihood2.40

showc2.39

oreal2.38

orry2.37

krit2.34

merce2.29

swers2.27

htaking2.26

Negative logits

thereafter-2.42

Tokens-2.32

}.-2.10

afterward-2.10

asel-2.07

Discussion-2.06

eventual-2.03

arsity-2.03

afterwards-2.00

1970-1.96

emed-1.94

shapeshifter-1.93

Eventually-1.93

FFFF-1.93

later-1.92

subsequent-1.91

};-1.88

[_-1.87

�士-1.86

bearing-1.85

INTERVAL 1.991 - 2.212
CONTAINS 0.000%

<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
going
Token going
Feature activation+0.000
forward
Token forward
Feature activation+0.000
."
Token."
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
Dead
TokenDead
Feature activation+2.027
hungry
Token hungry
Feature activation+1.348
:
Token:
Feature activation+0.712
Mother
Token Mother
Feature activation+1.295
-
Token-
Feature activation+0.548
to
Tokento
Feature activation+0.668
m
Tokenm
Feature activation+0.000
gonna
Token gonna
Feature activation+0.000
call
Token call
Feature activation+0.000
Ted
Token Ted
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.173
Posted
TokenPosted
Feature activation+2.212
in
Token in
Feature activation+1.065
Ċ
TokenĊ
Feature activation+0.300
Ċ
TokenĊ
Feature activation+0.062
Dear
TokenDear
Feature activation+0.920
friends
Token friends
Feature activation+0.678

INTERVAL 1.769 - 1.991
CONTAINS 0.002%

with
Token with
Feature activation+0.000
less
Token less
Feature activation+0.000
power
Token power
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.002
Sh
TokenSh
Feature activation+1.886
ining
Tokenining
Feature activation+1.789
Reson
Token Reson
Feature activation+1.401
ance
Tokenance
Feature activation+1.242
Re
Token Re
Feature activation+1.332
:
Token:
Feature activation+0.653
f
Tokenf
Feature activation+0.580
izers
Tokenizers
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ŀ
TokenĿ
Feature activation+0.000
âĢĵ
Token âĢĵ
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
Due
TokenDue
Feature activation+1.782
to
Token to
Feature activation+0.854
boring
Token boring
Feature activation+1.211
circumstances
Token circumstances
Feature activation+0.981
beyond
Token beyond
Feature activation+0.628
my
Token my
Feature activation+0.767
com
Tokencom
Feature activation+0.000
has
Token has
Feature activation+0.000
been
Token been
Feature activation+0.000
occupied
Token occupied
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
Pre
TokenPre
Feature activation+1.778
face
Tokenface
Feature activation+1.256
The
Token The
Feature activation+0.887
essence
Token essence
Feature activation+1.026
of
Token of
Feature activation+0.181
the
Token the
Feature activation+0.111
band
Token band
Feature activation+0.000
ing
Tokening
Feature activation+0.000
together
Token together
Feature activation+0.000
,
Token,
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
After
TokenAfter
Feature activation+1.861
talking
Token talking
Feature activation+1.501
about
Token about
Feature activation+1.069
how
Token how
Feature activation+0.830
unlikely
Token unlikely
Feature activation+0.853
it
Token it
Feature activation+0.618
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
the
Token the
Feature activation+0.000
Lab
Token Lab
Feature activation+0.000
.
Token.
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
Half
TokenHalf
Feature activation+1.775
of
Token of
Feature activation+0.899
the
Token the
Feature activation+0.775
residents
Token residents
Feature activation+0.975
of
Token of
Feature activation+0.482
Illinois
Token Illinois
Feature activation+0.755

INTERVAL 1.548 - 1.769
CONTAINS 0.005%

Like
Token Like
Feature activation+0.000
Loading
Token Loading
Feature activation+0.000
...
Token...
Feature activation+0.000
Related
Token Related
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
5
Token5
Feature activation+1.763
.
Token.
Feature activation+0.449
0
Token0
Feature activation+1.309
âĺħ
Token âĺħ
Feature activation+0.891
âĺħâĺħ
Tokenâĺħâĺħ
Feature activation+0.527
âĺħâĺħ
Tokenâĺħâĺħ
Feature activation+0.472
0000
Token0000
Feature activation+0.000
66
Token66
Feature activation+0.000
164
Token164
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
D
TokenD
Feature activation+1.488
rew
Tokenrew
Feature activation+1.597
In
Token In
Feature activation+0.970
ge
Tokenge
Feature activation+0.924
had
Token had
Feature activation+0.605
his
Token his
Feature activation+0.435
fingers
Token fingers
Feature activation+0.248
ľ
Tokenľ
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.045
Ly
TokenLy
Feature activation+1.952
ons
Tokenons
Feature activation+1.444
(
Token (
Feature activation+0.974
Photo
TokenPhoto
Feature activation+1.580
:
Token:
Feature activation+0.797
Facebook
Token Facebook
Feature activation+1.036
)
Token)
Feature activation+0.773
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Post
Token Post
Feature activation+0.000
and
Token and
Feature activation+0.000
Support
Token Support
Feature activation+0.000
:
Token:
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
3
Token3
Feature activation+1.560
)
Token)
Feature activation+0.882
How
Token How
Feature activation+1.109
"
Token "
Feature activation+0.667
false
Tokenfalse
Feature activation+0.479
equival
Token equival
Feature activation+0.301
before
Token before
Feature activation+0.000
she
Token she
Feature activation+0.000
disappeared
Token disappeared
Feature activation+0.000
.
Token.
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
We
TokenWe
Feature activation+1.623
suggest
Token suggest
Feature activation+1.160
that
Token that
Feature activation+0.964
signal
Token signal
Feature activation+1.112
convergence
Token convergence
Feature activation+0.771
between
Token between
Feature activation+0.475

INTERVAL 1.327 - 1.548
CONTAINS 0.009%

going
Token going
Feature activation+0.000
forward
Token forward
Feature activation+0.000
."
Token."
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
Dead
TokenDead
Feature activation+2.027
hungry
Token hungry
Feature activation+1.348
:
Token:
Feature activation+0.712
Mother
Token Mother
Feature activation+1.295
-
Token-
Feature activation+0.548
to
Tokento
Feature activation+0.668
-
Token-
Feature activation+0.343
ane
Tokenane
Feature activation+0.000
z
Tokenz
Feature activation+0.000
@
Token@
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.104
H
TokenH
Feature activation+1.742
ollywood
Tokenollywood
Feature activation+1.411
's
Token's
Feature activation+1.023
highest
Token highest
Feature activation+1.018
profile
Token profile
Feature activation+0.734
feminist
Token feminist
Feature activation+0.580
gets
Token gets
Feature activation+0.392
the
Token the
Feature activation+0.000
health
Token health
Feature activation+0.000
department
Token department
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
Hor
TokenHor
Feature activation+1.717
ace
Tokenace
Feature activation+1.525
Augustus
Token Augustus
Feature activation+1.071
Curtis
Token Curtis
Feature activation+1.129
VC
Token VC
Feature activation+0.904
(
Token (
Feature activation+0.424
7
Token7
Feature activation+0.569
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
READ
Token READ
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.101
Nich
TokenNich
Feature activation+1.835
olas
Tokenolas
Feature activation+1.708
Krist
Token Krist
Feature activation+1.390
of
Tokenof
Feature activation+1.167
and
Token and
Feature activation+0.602
Daniel
Token Daniel
Feature activation+1.033
Patrick
Token Patrick
Feature activation+0.981
Moy
Token Moy
Feature activation+0.651
start
Token start
Feature activation+0.000
enforcing
Token enforcing
Feature activation+0.000
laws
Token laws
Feature activation+0.000
.
Token.
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
Eight
TokenEight
Feature activation+1.450
months
Token months
Feature activation+0.617
ago
Token ago
Feature activation+0.394
,
Token,
Feature activation+0.000
Jeff
Token Jeff
Feature activation+0.156
G
Token G
Feature activation+0.000

INTERVAL 1.106 - 1.327
CONTAINS 0.010%

eman
Tokeneman
Feature activation+0.000
AP
TokenAP
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.138
Air
TokenAir
Feature activation+1.733
guns
Token guns
Feature activation+1.376
used
Token used
Feature activation+1.235
for
Token for
Feature activation+0.697
marine
Token marine
Feature activation+0.881
oil
Token oil
Feature activation+0.563
and
Token and
Feature activation+0.112
gas
Token gas
Feature activation+0.230
20
Token20
Feature activation+0.000
pm
Tokenpm
Feature activation+0.000
.
Token.
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
The
TokenThe
Feature activation+1.465
University
Token University
Feature activation+1.124
of
Token of
Feature activation+0.545
Virginia
Token Virginia
Feature activation+0.906
recently
Token recently
Feature activation+0.631
received
Token received
Feature activation+0.517
a
Token a
Feature activation+0.000
Wednesday
Token Wednesday
Feature activation+0.000
.
Token.
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
A
TokenA
Feature activation+1.662
team
Token team
Feature activation+1.430
led
Token led
Feature activation+1.269
by
Token by
Feature activation+0.815
post
Token post
Feature activation+0.938
doctoral
Tokendoctoral
Feature activation+0.747
associate
Token associate
Feature activation+0.591
John
Token John
Feature activation+0.189
Public
Token Public
Feature activation+0.000
Affairs
Token Affairs
Feature activation+0.000
.
Token.
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
I
TokenI
Feature activation+1.518
got
Token got
Feature activation+1.312
my
Token my
Feature activation+1.175
package
Token package
Feature activation+0.992
today
Token today
Feature activation+1.003
and
Token and
Feature activation+0.353
let
Token let
Feature activation+0.616
READ
Token READ
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.101
Nich
TokenNich
Feature activation+1.835
olas
Tokenolas
Feature activation+1.708
Krist
Token Krist
Feature activation+1.390
of
Tokenof
Feature activation+1.167
and
Token and
Feature activation+0.602
Daniel
Token Daniel
Feature activation+1.033
Patrick
Token Patrick
Feature activation+0.981
Moy
Token Moy
Feature activation+0.651
nih
Tokennih
Feature activation+0.216

INTERVAL 0.885 - 1.106
CONTAINS 0.014%

and
Token and
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
Q
TokenQ
Feature activation+1.529
atar
Tokenatar
Feature activation+1.169
are
Token are
Feature activation+0.897
set
Token set
Feature activation+1.010
to
Token to
Feature activation+0.238
offer
Token offer
Feature activation+0.621
£
Token £
Feature activation+0.427
175
Token175
Feature activation+0.133
m
Tokenm
Feature activation+0.000
Morgan
Token Morgan
Feature activation+0.000
said
Token said
Feature activation+0.000
.
Token.
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
>>
Token>>
Feature activation+1.608
K
TokenK
Feature activation+1.067
EEP
TokenEEP
Feature activation+0.918
GO
Token GO
Feature activation+0.832
ING
TokenING
Feature activation+0.621
TO
Token TO
Feature activation+0.351
SEE
Token SEE
Feature activation+0.351
to
Token to
Feature activation+0.000
hearing
Token hearing
Feature activation+0.000
.
Token.
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
Many
TokenMany
Feature activation+1.507
of
Token of
Feature activation+0.909
you
Token you
Feature activation+1.144
will
Token will
Feature activation+0.786
still
Token still
Feature activation+0.829
be
Token be
Feature activation+0.365
scratching
Token scratching
Feature activation+0.244
Police
TokenPolice
Feature activation+1.666
snap
Token snap
Feature activation+1.390
up
Token up
Feature activation+0.943
mud
Token mud
Feature activation+1.125
crab
Token crab
Feature activation+0.616
thought
Token thought
Feature activation+0.990
to
Token to
Feature activation+0.084
be
Token be
Feature activation+0.226
intruder
Token intruder
Feature activation+0.070
in
Token in
Feature activation+0.000
West
Token West
Feature activation+0.000
When
TokenWhen
Feature activation+1.799
it
Token it
Feature activation+1.248
comes
Token comes
Feature activation+1.359
to
Token to
Feature activation+0.837
climbing
Token climbing
Feature activation+1.173
water
Token water
Feature activation+0.896
falls
Tokenfalls
Feature activation+0.586
,
Token,
Feature activation+0.020
the
Token the
Feature activation+0.060
N
Token N
Feature activation+0.341
op
Tokenop
Feature activation+0.361

INTERVAL 0.664 - 0.885
CONTAINS 0.015%

<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
Anthony
TokenAnthony
Feature activation+1.581
Log
Token Log
Feature activation+1.525
istics
Tokenistics
Feature activation+1.132
for
Token for
Feature activation+0.688
Men
Token Men
Feature activation+0.772
shaving
Token shaving
Feature activation+0.721
cream
Token cream
Feature activation+0.639
is
Token is
Feature activation+0.171
the
Token the
Feature activation+0.000
first
Token first
Feature activation+0.127
<|endoftext|>
Token<|endoftext|>
Feature activation+0.034
One
TokenOne
Feature activation+1.724
New
Token New
Feature activation+1.433
York
Token York
Feature activation+1.200
City
Token City
Feature activation+0.845
Uber
Token Uber
Feature activation+0.859
driver
Token driver
Feature activation+0.525
made
Token made
Feature activation+0.714
his
Token his
Feature activation+0.378
tax
Token tax
Feature activation+0.283
filings
Token filings
Feature activation+0.393
of
Token of
Feature activation+0.000
8
Token 8
Feature activation+0.000
C
Token C
Feature activation+0.000
.
Token.
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
The
TokenThe
Feature activation+0.700
FBI
Token FBI
Feature activation+0.440
is
Token is
Feature activation+0.039
indeed
Token indeed
Feature activation+0.000
interested
Token interested
Feature activation+0.000
in
Token in
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.045
Ly
TokenLy
Feature activation+1.952
ons
Tokenons
Feature activation+1.444
(
Token (
Feature activation+0.974
Photo
TokenPhoto
Feature activation+1.580
:
Token:
Feature activation+0.797
Facebook
Token Facebook
Feature activation+1.036
)
Token)
Feature activation+0.773
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
A
TokenA
Feature activation+0.506
example
Token example
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
Looking
TokenLooking
Feature activation+1.471
for
Token for
Feature activation+0.666
news
Token news
Feature activation+0.702
you
Token you
Feature activation+0.660
can
Token can
Feature activation+0.291
trust
Token trust
Feature activation+0.093
?
Token?
Feature activation+0.000

INTERVAL 0.442 - 0.664
CONTAINS 0.019%

security
Token security
Feature activation+0.000
forces
Token forces
Feature activation+0.000
.
Token.
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
One
TokenOne
Feature activation+1.253
of
Token of
Feature activation+0.461
the
Token the
Feature activation+0.292
problems
Token problems
Feature activation+0.141
that
Token that
Feature activation+0.024
has
Token has
Feature activation+0.000
dogged
Token dogged
Feature activation+0.000
oma
Tokenoma
Feature activation+1.186
have
Token have
Feature activation+0.710
confirmed
Token confirmed
Feature activation+0.972
they
Token they
Feature activation+0.493
have
Token have
Feature activation+0.218
accepted
Token accepted
Feature activation+0.512
the
Token the
Feature activation+0.000
resignation
Token resignation
Feature activation+0.000
of
Token of
Feature activation+0.000
Cl
Token Cl
Feature activation+0.000
audio
Tokenaudio
Feature activation+0.000
.
Token.
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
With
TokenWith
Feature activation+1.482
an
Token an
Feature activation+0.937
impressive
Token impressive
Feature activation+0.784
group
Token group
Feature activation+0.587
of
Token of
Feature activation+0.095
future
Token future
Feature activation+0.354
television
Token television
Feature activation+0.122
and
Token and
Feature activation+0.000
movie
Token movie
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ŀ
TokenĿ
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
By
TokenBy
Feature activation+1.500
Ellen
Token Ellen
Feature activation+0.792
As
Token As
Feature activation+0.472
er
Tokener
Feature activation+0.205
me
Tokenme
Feature activation+0.402
ly
Tokenly
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
Police
TokenPolice
Feature activation+1.666
snap
Token snap
Feature activation+1.390
up
Token up
Feature activation+0.943
mud
Token mud
Feature activation+1.125
crab
Token crab
Feature activation+0.616
thought
Token thought
Feature activation+0.990
to
Token to
Feature activation+0.084
be
Token be
Feature activation+0.226
intruder
Token intruder
Feature activation+0.070
in
Token in
Feature activation+0.000

INTERVAL 0.221 - 0.442
CONTAINS 0.024%

.
Token.
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
Democratic
TokenDemocratic
Feature activation+1.177
National
Token National
Feature activation+0.573
Committee
Token Committee
Feature activation+0.353
staffer
Token staffer
Feature activation+0.429
Seth
Token Seth
Feature activation+0.247
Rich
Token Rich
Feature activation+0.008
,
Token,
Feature activation+0.000
who
Token who
Feature activation+0.000
was
Token was
Feature activation+0.000
highest
Token highest
Feature activation+1.018
profile
Token profile
Feature activation+0.734
feminist
Token feminist
Feature activation+0.580
gets
Token gets
Feature activation+0.392
sm
Token sm
Feature activation+0.227
acked
Tokenacked
Feature activation+0.323
around
Token around
Feature activation+0.000
by
Token by
Feature activation+0.000
film
Token film
Feature activation+0.000
critics
Token critics
Feature activation+0.000
for
Token for
Feature activation+0.000
cover
Token cover
Feature activation+0.000
.
Token.
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
BE
TokenBE
Feature activation+1.303
IJ
TokenIJ
Feature activation+0.580
ING
TokenING
Feature activation+0.333
(
Token (
Feature activation+0.227
Reuters
TokenReuters
Feature activation+0.408
)
Token)
Feature activation+0.000
-
Token -
Feature activation+0.000
China
Token China
Feature activation+0.000
is
Token is
Feature activation+0.128
an
Token an
Feature activation+0.232
easy
Token easy
Feature activation+0.304
week
Token week
Feature activation+0.082
night
Tokennight
Feature activation+0.117
dinner
Token dinner
Feature activation+0.335
fix
Token fix
Feature activation+0.000
.
Token.
Feature activation+0.000
This
Token This
Feature activation+0.000
chicken
Token chicken
Feature activation+0.000
curry
Token curry
Feature activation+0.000
Glenn
Token Glenn
Feature activation+1.264
Vol
Token Vol
Feature activation+0.967
iva
Tokeniva
Feature activation+0.868
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Wil
TokenWil
Feature activation+0.267
bur
Tokenbur
Feature activation+0.019
Glenn
Token Glenn
Feature activation+0.126
Vol
Token Vol
Feature activation+0.000
iva
Tokeniva
Feature activation+0.000
(
Token (
Feature activation+0.000

INTERVAL 0.000 - 0.221
CONTAINS 99.903%

way
Tokenway
Feature activation+0.000
race
Token race
Feature activation+0.000
led
Token led
Feature activation+0.000
by
Token by
Feature activation+0.000
Chow
Token Chow
Feature activation+0.000
with
Token with
Feature activation+0.000
36
Token 36
Feature activation+0.000
per
Token per
Feature activation+0.000
cent
Token cent
Feature activation+0.000
,
Token,
Feature activation+0.000
Ford
Token Ford
Feature activation+0.000
FCC
Token FCC
Feature activation+0.000
during
Token during
Feature activation+0.000
his
Token his
Feature activation+0.000
tenure
Token tenure
Feature activation+0.000
.
Token.
Feature activation+0.000
He
Token He
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
s
Tokens
Feature activation+0.000
never
Token never
Feature activation+0.000
met
Token met
Feature activation+0.000
votes
Token votes
Feature activation+0.000
)
Token)
Feature activation+0.000
Sh
Token Sh
Feature activation+0.000
ih
Tokenih
Feature activation+0.000
o
Tokeno
Feature activation+0.000
Suz
Token Suz
Feature activation+0.000
ui
Tokenui
Feature activation+0.000
(
Token (
Feature activation+0.000
146
Token146
Feature activation+0.000
votes
Token votes
Feature activation+0.000
)
Token)
Feature activation+0.000
since
Token since
Feature activation+0.000
1976
Token 1976
Feature activation+0.000
,
Token,
Feature activation+0.000
when
Token when
Feature activation+0.000
satellite
Token satellite
Feature activation+0.000
measurements
Token measurements
Feature activation+0.000
were
Token were
Feature activation+0.000
first
Token first
Feature activation+0.000
available
Token available
Feature activation+0.000
.
Token.
Feature activation+0.000
The
Token The
Feature activation+0.000
efforts
Token efforts
Feature activation+0.000
for
Token for
Feature activation+0.000
Linux
Token Linux
Feature activation+0.000
in
Token in
Feature activation+0.000
devices
Token devices
Feature activation+0.000
,
Token,
Feature activation+0.000
the
Token the
Feature activation+0.000
Xen
Token Xen
Feature activation+0.000
hyper
Token hyper
Feature activation+0.000
visor
Tokenvisor
Feature activation+0.000
,
Token,
Feature activation+0.000

BOTTOM ACTIVATIONS
MIN = 0.000

regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
anking
Tokenanking
Feature activation+0.000
the
Token the
Feature activation+0.000
sc
Token sc
Feature activation+0.000
iss
Tokeniss
Feature activation+0.000
ile
Tokenile
Feature activation+0.000
bonds
Token bonds
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
anking
Tokenanking
Feature activation+0.000
the
Token the
Feature activation+0.000
sc
Token sc
Feature activation+0.000
iss
Tokeniss
Feature activation+0.000
ile
Tokenile
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
anking
Tokenanking
Feature activation+0.000
the
Token the
Feature activation+0.000
sc
Token sc
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
anking
Tokenanking
Feature activation+0.000
the
Token the
Feature activation+0.000
sc
Token sc
Feature activation+0.000
iss
Tokeniss
Feature activation+0.000
ile
Tokenile
Feature activation+0.000
(
Token (
Feature activation+0.000
Michael
TokenMichael
Feature activation+0.000
C
Token C
Feature activation+0.000
aine
Tokenaine
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
ile
Tokenile
Feature activation+0.000
(
Token (
Feature activation+0.000
Michael
TokenMichael
Feature activation+0.000
C
Token C
Feature activation+0.000
aine
Tokenaine
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
(
Token (
Feature activation+0.000
Michael
TokenMichael
Feature activation+0.000
C
Token C
Feature activation+0.000
aine
Tokenaine
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Michael
TokenMichael
Feature activation+0.000
C
Token C
Feature activation+0.000
aine
Tokenaine
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
anking
Tokenanking
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
anking
Tokenanking
Feature activation+0.000
the
Token the
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
C
Token C
Feature activation+0.000
aine
Tokenaine
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
aine
Tokenaine
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000

Top feature 1 in H0.11: (feature 7517

TOP ACTIVATIONS
MAX = 1.674

<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
St
Token St
Feature activation+0.000
ake
Tokenake
Feature activation+0.430
in
Token in
Feature activation+0.000
Uran
Token Uran
Feature activation+0.000
ium
Tokenium
Feature activation+0.877
One
Token One
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Polit
TokenPolit
Feature activation+0.000
ifact
Tokenifact
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
.
Token.
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ŀ
TokenĿ
Feature activation+0.000
Tong
Token Tong
Feature activation+0.000
ue
Tokenue
Feature activation+0.853
-
Token-
Feature activation+0.000
in
Tokenin
Feature activation+0.000
-
Token-
Feature activation+0.000
che
Tokenche
Feature activation+0.000
ek
Tokenek
Feature activation+0.000
director
Tokendirector
Feature activation+0.000
Peter
Token Peter
Feature activation+0.091
Nicholson
Token Nicholson
Feature activation+0.251
and
Token and
Feature activation+0.000
Isabel
Token Isabel
Feature activation+0.010
le
Tokenle
Feature activation+0.839
Grey
Token Grey
Feature activation+0.000
,
Token,
Feature activation+0.000
Dart
Token Dart
Feature activation+0.000
m
Tokenm
Feature activation+0.000
oor
Tokenoor
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
Fried
Token Fried
Feature activation+0.000
Turkey
Token Turkey
Feature activation+0.000
Sand
Token Sand
Feature activation+0.000
wic
Tokenwic
Feature activation+0.561
hes
Tokenhes
Feature activation+0.831
are
Token are
Feature activation+0.000
available
Token available
Feature activation+0.000
in
Token in
Feature activation+0.000
two
Token two
Feature activation+0.000
varieties
Token varieties
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
credit
Token credit
Feature activation+0.000
:
Token:
Feature activation+0.000
i
Token i
Feature activation+0.000
Gam
TokenGam
Feature activation+0.000
ers
Tokeners
Feature activation+0.700
Youtube
Token Youtube
Feature activation+0.000
Channel
Token Channel
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
Tw
TokenTw
Feature activation+0.000
elve
Tokenelve
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
stole
Token stole
Feature activation+0.000
the
Token the
Feature activation+0.000
Delta
Token Delta
Feature activation+0.000
Fly
Token Fly
Feature activation+0.000
er
Tokener
Feature activation+0.699
II
Token II
Feature activation+0.000
and
Token and
Feature activation+0.000
took
Token took
Feature activation+0.000
I
Token I
Feature activation+0.000
che
Tokenche
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
ique
Tokenique
Feature activation+0.000
Maurit
Token Maurit
Feature activation+0.523
ania
Tokenania
Feature activation+1.004
Maurit
Token Maurit
Feature activation+0.150
ius
Tokenius
Feature activation+0.687
May
Token May
Feature activation+0.000
otte
Tokenotte
Feature activation+0.000
Mexico
Token Mexico
Feature activation+0.000
Mid
Token Mid
Feature activation+0.000
way
Tokenway
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
FOX
Token FOX
Feature activation+0.000
Sports
Token Sports
Feature activation+0.415
Exec
Token Exec
Feature activation+0.062
ut
Tokenut
Feature activation+0.381
ives
Tokenives
Feature activation+0.675
and
Token and
Feature activation+0.000
Pro
Token Pro
Feature activation+0.000
ducers
Tokenducers
Feature activation+0.000
using
Token using
Feature activation+0.000
Samsung
Token Samsung
Feature activation+0.000
READ
Token READ
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
Nich
TokenNich
Feature activation+0.000
olas
Tokenolas
Feature activation+0.293
Krist
Token Krist
Feature activation+0.175
of
Tokenof
Feature activation+0.657
and
Token and
Feature activation+0.000
Daniel
Token Daniel
Feature activation+0.000
Patrick
Token Patrick
Feature activation+0.000
Moy
Token Moy
Feature activation+0.000
nih
Tokennih
Feature activation+0.092
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
s
Tokens
Feature activation+0.000
Narc
Token Narc
Feature activation+0.000
otic
Tokenotic
Feature activation+0.634
Drug
Token Drug
Feature activation+0.000
Study
Token Study
Feature activation+0.000
Commission
Token Commission
Feature activation+0.000
called
Token called
Feature activation+0.000
LSD
Token LSD
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
soldiers
Token soldiers
Feature activation+0.000
and
Token and
Feature activation+0.000
ze
Token ze
Feature activation+0.000
ppel
Tokenppel
Feature activation+0.075
ins
Tokenins
Feature activation+0.600
to
Token to
Feature activation+0.000
claim
Token claim
Feature activation+0.000
an
Token an
Feature activation+0.000
inevitably
Token inevitably
Feature activation+0.000
short
Token short
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
north
Token north
Feature activation+0.000
of
Token of
Feature activation+0.000
Bur
Token Bur
Feature activation+0.000
und
Tokenund
Feature activation+0.043
i
Tokeni
Feature activation+0.595
's
Token's
Feature activation+0.000
capital
Token capital
Feature activation+0.000
Bu
Token Bu
Feature activation+0.000
j
Tokenj
Feature activation+0.000
umb
Tokenumb
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
planets
Token planets
Feature activation+0.000
and
Token and
Feature activation+0.000
Tra
Token Tra
Feature activation+0.000
pp
Tokenpp
Feature activation+0.151
ist
Tokenist
Feature activation+0.588
-
Token-
Feature activation+0.000
1
Token1
Feature activation+0.000
itself
Token itself
Feature activation+0.000
,
Token,
Feature activation+0.000
every
Token every
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
team
Token team
Feature activation+0.000
,
Token,
Feature activation+0.000
encouraging
Token encouraging
Feature activation+0.000
Ot
Token Ot
Feature activation+0.000
ters
Tokenters
Feature activation+0.585
to
Token to
Feature activation+0.000
commit
Token commit
Feature activation+0.000
to
Token to
Feature activation+0.000
a
Token a
Feature activation+0.000
shield
Token shield
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
need
Token need
Feature activation+0.000
a
Token a
Feature activation+0.000
TV
Token TV
Feature activation+0.000
Lic
Token Lic
Feature activation+0.000
ence
Tokenence
Feature activation+0.568
?
Token?
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Watch
TokenWatch
Feature activation+0.000
ing
Tokening
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
for
Token for
Feature activation+0.000
Ryan
Token Ryan
Feature activation+0.000
Tan
Token Tan
Feature activation+0.341
ne
Tokenne
Feature activation+0.453
hill
Tokenhill
Feature activation+0.561
on
Token on
Feature activation+0.000
the
Token the
Feature activation+0.000
Miami
Token Miami
Feature activation+0.000
Dolphins
Token Dolphins
Feature activation+0.000
and
Token and
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
it
Token it
Feature activation+0.000
and
Token and
Feature activation+0.000
a
Token a
Feature activation+0.000
Soci
Token Soci
Feature activation+0.000
ally
Tokenally
Feature activation+0.560
Aw
Token Aw
Feature activation+0.000
kward
Tokenkward
Feature activation+0.000
Penguin
Token Penguin
Feature activation+0.000
can
Token can
Feature activation+0.000
scream
Token scream
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
Robert
Token Robert
Feature activation+0.000
Baldwin
Token Baldwin
Feature activation+0.139
,
Token,
Feature activation+0.000
Heart
Token Heart
Feature activation+0.000
land
Tokenland
Feature activation+0.554
's
Token's
Feature activation+0.000
president
Token president
Feature activation+0.000
and
Token and
Feature activation+0.000
C
Token C
Feature activation+0.000
FO
TokenFO
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
Liberal
Token Liberal
Feature activation+0.000
Senator
Token Senator
Feature activation+0.000
Mac
Token Mac
Feature activation+0.000
Har
Token Har
Feature activation+0.000
b
Tokenb
Feature activation+0.553
who
Token who
Feature activation+0.000
has
Token has
Feature activation+0.000
lived
Token lived
Feature activation+0.000
in
Token in
Feature activation+0.000
Ottawa
Token Ottawa
Feature activation+0.000
said
Token said
Feature activation+0.000
"
Token "
Feature activation+0.000
Draw
TokenDraw
Feature activation+0.000
Superman
Token Superman
Feature activation+0.000
flex
Token flex
Feature activation+0.000
ing
Tokening
Feature activation+0.550
and
Token and
Feature activation+0.000
showing
Token showing
Feature activation+0.000
off
Token off
Feature activation+0.000
his
Token his
Feature activation+0.000
tattoos
Token tattoos
Feature activation+0.000

Top DFA by src position
MAX = 1.241

<|endoftext|>
Token<|endoftext|>
Feature activation+0.901
Top resid features:
St
Token St
Feature activation+0.068
Top resid features:
ake
Tokenake
Feature activation-0.034
Top resid features:
in
Token in
Feature activation-0.130
Top resid features:
Uran
Token Uran
Feature activation+0.902
Top resid features:
ium
Tokenium
Feature activation+0.376
Top resid features:
One
Token One
Feature activation+0.000
Top resid features:
Ċ
TokenĊ
Feature activation+0.000
Top resid features:
Ċ
TokenĊ
Feature activation+0.000
Top resid features:
Polit
TokenPolit
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.981
Top resid features:
.
Token.
Feature activation+0.111
Top resid features:
âĢ
TokenâĢ
Feature activation-0.024
Top resid features:
Ŀ
TokenĿ
Feature activation-0.081
Top resid features:
Tong
Token Tong
Feature activation+0.804
Top resid features:
ue
Tokenue
Feature activation+0.268
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.910
Top resid features:
director
Tokendirector
Feature activation+0.087
Top resid features:
Peter
Token Peter
Feature activation+0.094
Top resid features:
Nicholson
Token Nicholson
Feature activation-0.030
Top resid features:
and
Token and
Feature activation-0.253
Top resid features:
Isabel
Token Isabel
Feature activation+0.812
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.939
Top resid features:
Fried
Token Fried
Feature activation+0.269
Top resid features:
Turkey
Token Turkey
Feature activation+0.019
Top resid features:
Sand
Token Sand
Feature activation+0.114
Top resid features:
wic
Tokenwic
Feature activation+0.330
Top resid features:
hes
Tokenhes
Feature activation+0.366
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+1.030
Top resid features:
credit
Token credit
Feature activation+0.036
Top resid features:
:
Token:
Feature activation-0.047
Top resid features:
i
Token i
Feature activation-0.064
Top resid features:
Gam
TokenGam
Feature activation+0.538
Top resid features:
ers
Tokeners
Feature activation+0.412
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+1.069
Top resid features:
stole
Token stole
Feature activation+0.039
Top resid features:
the
Token the
Feature activation-0.155
Top resid features:
Delta
Token Delta
Feature activation+0.002
Top resid features:
Fly
Token Fly
Feature activation+0.576
Top resid features:
er
Tokener
Feature activation+0.373
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+1.167
Top resid features:
ique
Tokenique
Feature activation+0.048
Top resid features:
Maurit
Token Maurit
Feature activation+0.190
Top resid features:
ania
Tokenania
Feature activation-0.078
Top resid features:
Maurit
Token Maurit
Feature activation+0.304
Top resid features:
ius
Tokenius
Feature activation+0.263
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+1.086
Top resid features:
FOX
Token FOX
Feature activation+0.151
Top resid features:
Sports
Token Sports
Feature activation-0.068
Top resid features:
Exec
Token Exec
Feature activation+0.582
Top resid features:
ut
Tokenut
Feature activation-0.052
Top resid features:
ives
Tokenives
Feature activation+0.182
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.762
Top resid features:
READ
Token READ
Feature activation+0.085
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.230
Top resid features:
Nich
TokenNich
Feature activation+0.133
Top resid features:
olas
Tokenolas
Feature activation+0.099
Top resid features:
Krist
Token Krist
Feature activation+0.349
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+1.121
Top resid features:
âĢ
TokenâĢ
Feature activation-0.012
Top resid features:
Ļ
TokenĻ
Feature activation-0.115
Top resid features:
s
Tokens
Feature activation-0.171
Top resid features:
Narc
Token Narc
Feature activation+0.922
Top resid features:
otic
Tokenotic
Feature activation+0.094
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.976
Top resid features:
soldiers
Token soldiers
Feature activation+0.025
Top resid features:
and
Token and
Feature activation-0.155
Top resid features:
ze
Token ze
Feature activation+0.272
Top resid features:
ppel
Tokenppel
Feature activation+0.320
Top resid features:
ins
Tokenins
Feature activation+0.368
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+1.006
Top resid features:
north
Token north
Feature activation+0.022
Top resid features:
of
Token of
Feature activation-0.052
Top resid features:
Bur
Token Bur
Feature activation+0.417
Top resid features:
und
Tokenund
Feature activation-0.022
Top resid features:
i
Tokeni
Feature activation+0.429
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+1.067
Top resid features:
planets
Token planets
Feature activation+0.050
Top resid features:
and
Token and
Feature activation-0.154
Top resid features:
Tra
Token Tra
Feature activation+0.556
Top resid features:
pp
Tokenpp
Feature activation-0.015
Top resid features:
ist
Tokenist
Feature activation+0.289
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+1.051
Top resid features:
team
Token team
Feature activation-0.035
Top resid features:
,
Token,
Feature activation-0.131
Top resid features:
encouraging
Token encouraging
Feature activation-0.137
Top resid features:
Ot
Token Ot
Feature activation+0.657
Top resid features:
ters
Tokenters
Feature activation+0.387
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+1.192
Top resid features:
need
Token need
Feature activation+0.017
Top resid features:
a
Token a
Feature activation-0.136
Top resid features:
TV
Token TV
Feature activation-0.031
Top resid features:
Lic
Token Lic
Feature activation+0.639
Top resid features:
ence
Tokenence
Feature activation+0.094
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+1.241
Top resid features:
for
Token for
Feature activation-0.087
Top resid features:
Ryan
Token Ryan
Feature activation+0.023
Top resid features:
Tan
Token Tan
Feature activation+0.431
Top resid features:
ne
Tokenne
Feature activation-0.023
Top resid features:
hill
Tokenhill
Feature activation+0.182
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+1.072
Top resid features:
it
Token it
Feature activation-0.074
Top resid features:
and
Token and
Feature activation-0.163
Top resid features:
a
Token a
Feature activation-0.164
Top resid features:
Soci
Token Soci
Feature activation+0.967
Top resid features:
ally
Tokenally
Feature activation+0.128
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+1.051
Top resid features:
Robert
Token Robert
Feature activation+0.158
Top resid features:
Baldwin
Token Baldwin
Feature activation+0.026
Top resid features:
,
Token,
Feature activation-0.105
Top resid features:
Heart
Token Heart
Feature activation+0.288
Top resid features:
land
Tokenland
Feature activation+0.341
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.973
Top resid features:
Liberal
Token Liberal
Feature activation+0.014
Top resid features:
Senator
Token Senator
Feature activation+0.026
Top resid features:
Mac
Token Mac
Feature activation+0.208
Top resid features:
Har
Token Har
Feature activation+0.270
Top resid features:
b
Tokenb
Feature activation+0.268
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.948
Top resid features:
said
Token said
Feature activation-0.013
Top resid features:
"
Token "
Feature activation-0.083
Top resid features:
Draw
TokenDraw
Feature activation+0.267
Top resid features:
Superman
Token Superman
Feature activation+0.021
Top resid features:
flex
Token flex
Feature activation+0.202
Top resid features:

Decoder Weights Distribution

Head 0: 0.06

Head 1: 0.09

Head 2: 0.07

Head 3: 0.09

Head 4: 0.04

Head 5: 0.04

Head 6: 0.07

Head 7: 0.06

Head 8: 0.07

Head 9: 0.17

Head 10: 0.08

Head 11: 0.15

Positive logits

lihood3.09

proble3.01

NetMessage2.93

aukee2.75

advertisement2.67

ktop2.67

hess2.61

pmwiki2.58

esson2.52

aston2.48

hedon2.46

hyde2.46

rahim2.46

ingred2.39

terday2.38

suspic2.38

haar2.38

wolves2.36

advoc2.34

perse2.33

Negative logits

CLSID-2.58

[|-2.36

Hilbert-2.24

��-2.19

Term-2.14

foundland-2.08

iversary-2.06

}.-2.06

utory-2.01

erva-1.90

[_-1.88

��-1.85

Tokens-1.83

thora-1.82

Conversation-1.80

occasion-1.79

default-1.79

borrowed-1.78

-1.78

amac-1.78

INTERVAL 1.507 - 1.674
CONTAINS 0.001%

INTERVAL 1.340 - 1.507
CONTAINS 0.002%

INTERVAL 1.172 - 1.340
CONTAINS 0.003%

INTERVAL 1.005 - 1.172
CONTAINS 0.006%

INTERVAL 0.837 - 1.005
CONTAINS 0.009%

director
Tokendirector
Feature activation+0.000
Peter
Token Peter
Feature activation+0.091
Nicholson
Token Nicholson
Feature activation+0.251
and
Token and
Feature activation+0.000
Isabel
Token Isabel
Feature activation+0.010
le
Tokenle
Feature activation+0.839
Grey
Token Grey
Feature activation+0.000
,
Token,
Feature activation+0.000
Dart
Token Dart
Feature activation+0.000
m
Tokenm
Feature activation+0.000
oor
Tokenoor
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
.
Token.
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ŀ
TokenĿ
Feature activation+0.000
Tong
Token Tong
Feature activation+0.000
ue
Tokenue
Feature activation+0.853
-
Token-
Feature activation+0.000
in
Tokenin
Feature activation+0.000
-
Token-
Feature activation+0.000
che
Tokenche
Feature activation+0.000
ek
Tokenek
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
St
Token St
Feature activation+0.000
ake
Tokenake
Feature activation+0.430
in
Token in
Feature activation+0.000
Uran
Token Uran
Feature activation+0.000
ium
Tokenium
Feature activation+0.877
One
Token One
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Polit
TokenPolit
Feature activation+0.000
ifact
Tokenifact
Feature activation+0.000

INTERVAL 0.670 - 0.837
CONTAINS 0.014%

<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
ique
Tokenique
Feature activation+0.000
Maurit
Token Maurit
Feature activation+0.523
ania
Tokenania
Feature activation+1.004
Maurit
Token Maurit
Feature activation+0.150
ius
Tokenius
Feature activation+0.687
May
Token May
Feature activation+0.000
otte
Tokenotte
Feature activation+0.000
Mexico
Token Mexico
Feature activation+0.000
Mid
Token Mid
Feature activation+0.000
way
Tokenway
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
FOX
Token FOX
Feature activation+0.000
Sports
Token Sports
Feature activation+0.415
Exec
Token Exec
Feature activation+0.062
ut
Tokenut
Feature activation+0.381
ives
Tokenives
Feature activation+0.675
and
Token and
Feature activation+0.000
Pro
Token Pro
Feature activation+0.000
ducers
Tokenducers
Feature activation+0.000
using
Token using
Feature activation+0.000
Samsung
Token Samsung
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
credit
Token credit
Feature activation+0.000
:
Token:
Feature activation+0.000
i
Token i
Feature activation+0.000
Gam
TokenGam
Feature activation+0.000
ers
Tokeners
Feature activation+0.700
Youtube
Token Youtube
Feature activation+0.000
Channel
Token Channel
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
Tw
TokenTw
Feature activation+0.000
elve
Tokenelve
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
stole
Token stole
Feature activation+0.000
the
Token the
Feature activation+0.000
Delta
Token Delta
Feature activation+0.000
Fly
Token Fly
Feature activation+0.000
er
Tokener
Feature activation+0.699
II
Token II
Feature activation+0.000
and
Token and
Feature activation+0.000
took
Token took
Feature activation+0.000
I
Token I
Feature activation+0.000
che
Tokenche
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
Fried
Token Fried
Feature activation+0.000
Turkey
Token Turkey
Feature activation+0.000
Sand
Token Sand
Feature activation+0.000
wic
Tokenwic
Feature activation+0.561
hes
Tokenhes
Feature activation+0.831
are
Token are
Feature activation+0.000
available
Token available
Feature activation+0.000
in
Token in
Feature activation+0.000
two
Token two
Feature activation+0.000
varieties
Token varieties
Feature activation+0.000

INTERVAL 0.502 - 0.670
CONTAINS 0.021%

READ
Token READ
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
Nich
TokenNich
Feature activation+0.000
olas
Tokenolas
Feature activation+0.293
Krist
Token Krist
Feature activation+0.175
of
Tokenof
Feature activation+0.657
and
Token and
Feature activation+0.000
Daniel
Token Daniel
Feature activation+0.000
Patrick
Token Patrick
Feature activation+0.000
Moy
Token Moy
Feature activation+0.000
nih
Tokennih
Feature activation+0.092
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
Robert
Token Robert
Feature activation+0.000
Baldwin
Token Baldwin
Feature activation+0.139
,
Token,
Feature activation+0.000
Heart
Token Heart
Feature activation+0.000
land
Tokenland
Feature activation+0.554
's
Token's
Feature activation+0.000
president
Token president
Feature activation+0.000
and
Token and
Feature activation+0.000
C
Token C
Feature activation+0.000
FO
TokenFO
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
HIV
Token HIV
Feature activation+0.000
Prevention
Token Prevention
Feature activation+0.329
:
Token:
Feature activation+0.000
Inf
Token Inf
Feature activation+0.000
ant
Tokenant
Feature activation+0.508
M
Token M
Feature activation+0.000
ogen
Tokenogen
Feature activation+0.000
clamp
Token clamp
Feature activation+0.000
alternatives
Token alternatives
Feature activation+0.000
,
Token,
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
planets
Token planets
Feature activation+0.000
and
Token and
Feature activation+0.000
Tra
Token Tra
Feature activation+0.000
pp
Tokenpp
Feature activation+0.151
ist
Tokenist
Feature activation+0.588
-
Token-
Feature activation+0.000
1
Token1
Feature activation+0.000
itself
Token itself
Feature activation+0.000
,
Token,
Feature activation+0.000
every
Token every
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
Teen
Token Teen
Feature activation+0.000
Che
Token Che
Feature activation+0.107
ez
Tokenez
Feature activation+0.148
Kab
Token Kab
Feature activation+0.000
hi
Tokenhi
Feature activation+0.511
Und
Token Und
Feature activation+0.000
erest
Tokenerest
Feature activation+0.000
imate
Tokenimate
Feature activation+0.000
Mat
Token Mat
Feature activation+0.000
K
Token K
Feature activation+0.000

INTERVAL 0.335 - 0.502
CONTAINS 0.034%

<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
as
Token as
Feature activation+0.000
a
Token a
Feature activation+0.000
Tex
Token Tex
Feature activation+0.000
-
Token-
Feature activation+0.000
Mex
TokenMex
Feature activation+0.423
restaurant
Token restaurant
Feature activation+0.000
but
Token but
Feature activation+0.000
you
Token you
Feature activation+0.000
'll
Token'll
Feature activation+0.000
find
Token find
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
ot
Tokenot
Feature activation+0.000
ini
Tokenini
Feature activation+0.087
,
Token,
Feature activation+0.000
Thr
Token Thr
Feature activation+0.000
ace
Tokenace
Feature activation+0.494
23
Token 23
Feature activation+0.000
rd
Tokenrd
Feature activation+0.000
Arm
Token Arm
Feature activation+0.000
oured
Tokenoured
Feature activation+0.000
Brigade
Token Brigade
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
quite
Token quite
Feature activation+0.000
possibly
Token possibly
Feature activation+0.066
,
Token,
Feature activation+0.000
haw
Token haw
Feature activation+0.000
ks
Tokenks
Feature activation+0.349
on
Token on
Feature activation+0.000
the
Token the
Feature activation+0.000
encryption
Token encryption
Feature activation+0.000
issue
Token issue
Feature activation+0.000
may
Token may
Feature activation+0.000
Attorney
Token Attorney
Feature activation+0.000
General
Token General
Feature activation+0.692
Hu
Token Hu
Feature activation+0.136
bert
Tokenbert
Feature activation+0.365
Humph
Token Humph
Feature activation+0.133
rey
Tokenrey
Feature activation+0.465
III
Token III
Feature activation+0.000
.
Token.
Feature activation+0.000
Enough
Token Enough
Feature activation+0.000
voters
Token voters
Feature activation+0.000
decided
Token decided
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
A
Token A
Feature activation+0.000
TV
Token TV
Feature activation+0.001
:
Token:
Feature activation+0.000
Kiw
Token Kiw
Feature activation+0.000
i
Tokeni
Feature activation+0.354
smart
Token smart
Feature activation+0.000
TV
Token TV
Feature activation+0.000
owners
Token owners
Feature activation+0.000
have
Token have
Feature activation+0.000
been
Token been
Feature activation+0.000

INTERVAL 0.167 - 0.335
CONTAINS 0.047%

<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
,
Token,
Feature activation+0.000
000
Token000
Feature activation+0.000
to
Token to
Feature activation+0.000
Adv
Token Adv
Feature activation+0.000
ancing
Tokenancing
Feature activation+0.221
Wisconsin
Token Wisconsin
Feature activation+0.000
,
Token ,
Feature activation+0.000
a
Token a
Feature activation+0.000
tax
Token tax
Feature activation+0.000
-
Token-
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
Ey
Token Ey
Feature activation+0.042
j
Tokenj
Feature activation+0.515
af
Tokenaf
Feature activation+0.228
j
Tokenj
Feature activation+0.303
all
Tokenall
Feature activation+0.221
aj
Tokenaj
Feature activation+0.000
ö
Tokenö
Feature activation+0.000
k
Tokenk
Feature activation+0.000
ull
Tokenull
Feature activation+0.000
in
Token in
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
Foreign
Token Foreign
Feature activation+0.000
Ministry
Token Ministry
Feature activation+0.700
spokesman
Token spokesman
Feature activation+0.052
Lu
Token Lu
Feature activation+0.000
Kang
Token Kang
Feature activation+0.200
reiterated
Token reiterated
Feature activation+0.000
China
Token China
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
s
Tokens
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
s
Tokens
Feature activation+0.000
or
Token or
Feature activation+0.000
mar
Token mar
Feature activation+0.000
ath
Tokenath
Feature activation+0.000
ons
Tokenons
Feature activation+0.170
as
Token as
Feature activation+0.000
my
Token my
Feature activation+0.000
schedule
Token schedule
Feature activation+0.000
would
Token would
Feature activation+0.000
allow
Token allow
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
Wichita
Token Wichita
Feature activation+0.000
State
Token State
Feature activation+0.970
:
Token:
Feature activation+0.000
Marc
Token Marc
Feature activation+0.000
in
Tokenin
Feature activation+0.181
G
Token G
Feature activation+0.000
ort
Tokenort
Feature activation+0.000
at
Tokenat
Feature activation+0.000
-
Token -
Feature activation+0.000
Guy
Token Guy
Feature activation+0.000

INTERVAL 0.000 - 0.167
CONTAINS 99.864%

be
Token be
Feature activation+0.000
pie
Token pie
Feature activation+0.000
-
Token-
Feature activation+0.000
eyed
Tokeneyed
Feature activation+0.000
days
Token days
Feature activation+0.000
before
Token before
Feature activation+0.000
a
Token a
Feature activation+0.000
match
Token match
Feature activation+0.000
during
Token during
Feature activation+0.000
a
Token a
Feature activation+0.000
season
Token season
Feature activation+0.000
!
Token!
Feature activation+0.000
Sign
Token Sign
Feature activation+0.000
up
Token up
Feature activation+0.000
for
Token for
Feature activation+0.000
more
Token more
Feature activation+0.000
newsletters
Token newsletters
Feature activation+0.000
here
Token here
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
We
TokenWe
Feature activation+0.000
were
Token were
Feature activation+0.000
remember
Token remember
Feature activation+0.000
Adam
Token Adam
Feature activation+0.000
Hunt
Token Hunt
Feature activation+0.000
,
Token,
Feature activation+0.000
right
Token right
Feature activation+0.000
?
Token?
Feature activation+0.000
Hunt
Token Hunt
Feature activation+0.000
was
Token was
Feature activation+0.000
the
Token the
Feature activation+0.000
first
Token first
Feature activation+0.000
person
Token person
Feature activation+0.000
Pharaoh
Token Pharaoh
Feature activation+0.000
victory
Token victory
Feature activation+0.000
Krist
Token Krist
Feature activation+0.000
ian
Tokenian
Feature activation+0.000
D
Token D
Feature activation+0.000
yer
Tokenyer
Feature activation+0.000
:
Token:
Feature activation+0.000
Is
Token Is
Feature activation+0.000
Triple
Token Triple
Feature activation+0.000
Crown
Token Crown
Feature activation+0.000
bad
Token bad
Feature activation+0.000
sole
Token sole
Feature activation+0.000
purpose
Token purpose
Feature activation+0.000
would
Token would
Feature activation+0.000
be
Token be
Feature activation+0.000
to
Token to
Feature activation+0.000
protect
Token protect
Feature activation+0.000
our
Token our
Feature activation+0.000
people
Token people
Feature activation+0.000
and
Token and
Feature activation+0.000
towns
Token towns
Feature activation+0.000
in
Token in
Feature activation+0.000

BOTTOM ACTIVATIONS
MIN = 0.000

regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
anking
Tokenanking
Feature activation+0.000
the
Token the
Feature activation+0.000
sc
Token sc
Feature activation+0.000
iss
Tokeniss
Feature activation+0.000
ile
Tokenile
Feature activation+0.000
bonds
Token bonds
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
anking
Tokenanking
Feature activation+0.000
the
Token the
Feature activation+0.000
sc
Token sc
Feature activation+0.000
iss
Tokeniss
Feature activation+0.000
ile
Tokenile
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
anking
Tokenanking
Feature activation+0.000
the
Token the
Feature activation+0.000
sc
Token sc
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
anking
Tokenanking
Feature activation+0.000
the
Token the
Feature activation+0.000
sc
Token sc
Feature activation+0.000
iss
Tokeniss
Feature activation+0.000
ile
Tokenile
Feature activation+0.000
(
Token (
Feature activation+0.000
Michael
TokenMichael
Feature activation+0.000
C
Token C
Feature activation+0.000
aine
Tokenaine
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
ile
Tokenile
Feature activation+0.000
(
Token (
Feature activation+0.000
Michael
TokenMichael
Feature activation+0.000
C
Token C
Feature activation+0.000
aine
Tokenaine
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
(
Token (
Feature activation+0.000
Michael
TokenMichael
Feature activation+0.000
C
Token C
Feature activation+0.000
aine
Tokenaine
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Michael
TokenMichael
Feature activation+0.000
C
Token C
Feature activation+0.000
aine
Tokenaine
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
anking
Tokenanking
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
anking
Tokenanking
Feature activation+0.000
the
Token the
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
C
Token C
Feature activation+0.000
aine
Tokenaine
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
aine
Tokenaine
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000

Top feature 2 in H0.11: (feature 19667

TOP ACTIVATIONS
MAX = 1.953

<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
LI
TokenLI
Feature activation+0.008
Los
Token Los
Feature activation+0.000
Angeles
Token Angeles
Feature activation+0.000
(@
Token (@
Feature activation+0.000
ul
Tokenul
Feature activation+0.686
il
Tokenil
Feature activation+0.376
os
Tokenos
Feature activation+0.129
angel
Tokenangel
Feature activation+0.000
es
Tokenes
Feature activation+0.000
)
Token)
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
g
Tokeng
Feature activation+0.171
.,
Token.,
Feature activation+0.000
B
Token B
Feature activation+0.000
ial
Tokenial
Feature activation+0.853
y
Tokeny
Feature activation+0.572
st
Tokenst
Feature activation+0.000
ok
Tokenok
Feature activation+0.000
,
Token,
Feature activation+0.000
2006
Token 2006
Feature activation+0.000
;
Token;
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
spike
Token spike
Feature activation+0.000
"
Token"
Feature activation+0.000
that
Token that
Feature activation+0.000
v
Token v
Feature activation+0.000
ented
Tokenented
Feature activation+0.569
from
Token from
Feature activation+0.000
the
Token the
Feature activation+0.000
subs
Token subs
Feature activation+0.000
ur
Tokenur
Feature activation+0.000
face
Tokenface
Feature activation+0.000
with
Token with
Feature activation+0.000
walk
Token walk
Feature activation+0.000
ie
Tokenie
Feature activation+0.833
-
Token-
Feature activation+0.000
talk
Tokentalk
Feature activation+0.000
ies
Tokenies
Feature activation+0.524
to
Token to
Feature activation+0.000
update
Token update
Feature activation+0.000
the
Token the
Feature activation+0.000
crew
Token crew
Feature activation+0.000
about
Token about
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
G
TokenG
Feature activation+0.000
RI
TokenRI
Feature activation+0.000
ZZ
TokenZZ
Feature activation+0.000
L
TokenL
Feature activation+0.000
IES
TokenIES
Feature activation+0.523
_
Token_
Feature activation+0.000
FB
TokenFB
Feature activation+0.000
)
Token)
Feature activation+0.000
became
Token became
Feature activation+0.000
home
Token home
Feature activation+0.000
itz
Tokenitz
Feature activation+0.465
man
Tokenman
Feature activation+0.000
,
Token,
Feature activation+0.000
the
Token the
Feature activation+0.000
W
Token W
Feature activation+0.000
ies
Tokenies
Feature activation+0.519
enthal
Tokenenthal
Feature activation+0.000
Center
Token Center
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
s
Tokens
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
featured
Token featured
Feature activation+0.000
two
Token two
Feature activation+0.000
stunning
Token stunning
Feature activation+0.000
ups
Token ups
Feature activation+0.000
ets
Tokenets
Feature activation+0.509
that
Token that
Feature activation+0.000
bolstered
Token bolstered
Feature activation+0.000
En
Token En
Feature activation+0.000
V
TokenV
Feature activation+0.000
y
Tokeny
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
born
Token born
Feature activation+0.000
in
Token in
Feature activation+0.000
L
Token L
Feature activation+0.000
le
Tokenle
Feature activation+0.000
ida
Tokenida
Feature activation+0.463
,
Token,
Feature activation+0.000
Spain
Token Spain
Feature activation+0.000
,
Token,
Feature activation+0.000
the
Token the
Feature activation+0.000
son
Token son
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
clearly
Token clearly
Feature activation+0.000
becoming
Token becoming
Feature activation+0.000
more
Token more
Feature activation+0.000
western
Token western
Feature activation+0.000
ized
Tokenized
Feature activation+0.428
at
Token at
Feature activation+0.000
a
Token a
Feature activation+0.000
rapid
Token rapid
Feature activation+0.000
pace
Token pace
Feature activation+0.000
.
Token.
Feature activation+0.000
http
Token http
Feature activation+0.000
://
Token://
Feature activation+0.000
www
Tokenwww
Feature activation+0.000
.
Token.
Feature activation+0.000
em
Tokenem
Feature activation+0.000
ily
Tokenily
Feature activation+0.405
review
Tokenreview
Feature activation+0.000
s
Tokens
Feature activation+0.000
.
Token.
Feature activation+0.000
com
Tokencom
Feature activation+0.000
/
Token/
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
especially
Token especially
Feature activation+0.000
since
Token since
Feature activation+0.000
Sergio
Token Sergio
Feature activation+0.000
March
Token March
Feature activation+0.000
ion
Tokenion
Feature activation+0.383
ne
Tokenne
Feature activation+0.000
has
Token has
Feature activation+0.000
stated
Token stated
Feature activation+0.000
that
Token that
Feature activation+0.000
it
Token it
Feature activation+0.000
LI
TokenLI
Feature activation+0.008
Los
Token Los
Feature activation+0.000
Angeles
Token Angeles
Feature activation+0.000
(@
Token (@
Feature activation+0.000
ul
Tokenul
Feature activation+0.686
il
Tokenil
Feature activation+0.376
os
Tokenos
Feature activation+0.129
angel
Tokenangel
Feature activation+0.000
es
Tokenes
Feature activation+0.000
)
Token)
Feature activation+0.000
January
Token January
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
based
Token based
Feature activation+0.000
on
Token on
Feature activation+0.000
Us
Token Us
Feature activation+0.090
en
Tokenen
Feature activation+0.000
et
Tokenet
Feature activation+0.355
that
Token that
Feature activation+0.000
started
Token started
Feature activation+0.000
in
Token in
Feature activation+0.000
1995
Token 1995
Feature activation+0.000
.
Token.
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
ier
Tokenier
Feature activation+1.655
Come
Token Come
Feature activation+0.000
to
Token to
Feature activation+0.000
Boulevard
Token Boulevard
Feature activation+0.000
ier
Tokenier
Feature activation+0.345
to
Token to
Feature activation+0.000
get
Token get
Feature activation+0.000
in
Token in
Feature activation+0.000
touch
Token touch
Feature activation+0.000
with
Token with
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
Civil
Token Civil
Feature activation+0.000
Aviation
Token Aviation
Feature activation+0.000
Organization
Token Organization
Feature activation+0.000
(
Token (
Feature activation+0.000
ICA
TokenICA
Feature activation+0.332
O
TokenO
Feature activation+0.000
)
Token)
Feature activation+0.000
audit
Token audit
Feature activation+0.000
that
Token that
Feature activation+0.000
it
Token it
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ŀ
TokenĿ
Feature activation+0.000
J
Token J
Feature activation+0.000
at
Tokenat
Feature activation+0.000
ia
Tokenia
Feature activation+0.328
said
Token said
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
In
TokenIn
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
football
Token football
Feature activation+0.000
legend
Token legend
Feature activation+0.000
Th
Token Th
Feature activation+0.000
ier
Tokenier
Feature activation+0.459
ry
Tokenry
Feature activation+0.307
Henry
Token Henry
Feature activation+0.000
,
Token,
Feature activation+0.000
C
Token C
Feature activation+0.000
ound
Tokenound
Feature activation+0.000
oul
Tokenoul
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
ai
Tokenai
Feature activation+1.078
yan
Tokenyan
Feature activation+0.421
super
Token super
Feature activation+0.000
-
Token-
Feature activation+0.000
ty
Tokenty
Feature activation+0.303
ph
Tokenph
Feature activation+0.000
oon
Tokenoon
Feature activation+0.000
.
Token.
Feature activation+0.000
I
Token I
Feature activation+0.000
'm
Token'm
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
remarks
Token remarks
Feature activation+0.000
,
Token,
Feature activation+0.000
Mr
Token Mr
Feature activation+0.000
Mans
Token Mans
Feature activation+0.000
our
Tokenour
Feature activation+0.298
said
Token said
Feature activation+0.000
he
Token he
Feature activation+0.000
had
Token had
Feature activation+0.000
secured
Token secured
Feature activation+0.000
the
Token the
Feature activation+0.000
,
Token,
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ŀ
TokenĿ
Feature activation+0.000
Men
Token Men
Feature activation+0.000
z
Tokenz
Feature activation+0.000
ies
Tokenies
Feature activation+0.296
said
Token said
Feature activation+0.000
.
Token.
Feature activation+0.000
âĢ
Token âĢ
Feature activation+0.000
ľ
Tokenľ
Feature activation+0.000
Whether
TokenWhether
Feature activation+0.000

Top DFA by src position
MAX = 2.617

<|endoftext|>
Token<|endoftext|>
Feature activation+1.140
Top resid features:
LI
TokenLI
Feature activation-0.030
Top resid features:
Los
Token Los
Feature activation+0.086
Top resid features:
Angeles
Token Angeles
Feature activation+0.127
Top resid features:
(@
Token (@
Feature activation+0.215
Top resid features:
ul
Tokenul
Feature activation+2.039
Top resid features:
il
Tokenil
Feature activation+0.000
Top resid features:
os
Tokenos
Feature activation+0.000
Top resid features:
angel
Tokenangel
Feature activation+0.000
Top resid features:
es
Tokenes
Feature activation+0.000
Top resid features:
)
Token)
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+1.240
Top resid features:
g
Tokeng
Feature activation-0.015
Top resid features:
.,
Token.,
Feature activation+0.029
Top resid features:
B
Token B
Feature activation-0.023
Top resid features:
ial
Tokenial
Feature activation+0.370
Top resid features:
y
Tokeny
Feature activation+1.861
Top resid features:
st
Tokenst
Feature activation+0.000
Top resid features:
ok
Tokenok
Feature activation+0.000
Top resid features:
,
Token,
Feature activation+0.000
Top resid features:
2006
Token 2006
Feature activation+0.000
Top resid features:
;
Token;
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+1.092
Top resid features:
spike
Token spike
Feature activation-0.008
Top resid features:
"
Token"
Feature activation-0.013
Top resid features:
that
Token that
Feature activation-0.025
Top resid features:
v
Token v
Feature activation+0.098
Top resid features:
ented
Tokenented
Feature activation+2.317
Top resid features:
from
Token from
Feature activation+0.000
Top resid features:
the
Token the
Feature activation+0.000
Top resid features:
subs
Token subs
Feature activation+0.000
Top resid features:
ur
Tokenur
Feature activation+0.000
Top resid features:
face
Tokenface
Feature activation+0.000
Top resid features:
with
Token with
Feature activation-0.044
Top resid features:
walk
Token walk
Feature activation-0.030
Top resid features:
ie
Tokenie
Feature activation+0.311
Top resid features:
-
Token-
Feature activation+0.007
Top resid features:
talk
Tokentalk
Feature activation+0.167
Top resid features:
ies
Tokenies
Feature activation+2.134
Top resid features:
to
Token to
Feature activation+0.000
Top resid features:
update
Token update
Feature activation+0.000
Top resid features:
the
Token the
Feature activation+0.000
Top resid features:
crew
Token crew
Feature activation+0.000
Top resid features:
about
Token about
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+1.181
Top resid features:
G
TokenG
Feature activation-0.071
Top resid features:
RI
TokenRI
Feature activation+0.109
Top resid features:
ZZ
TokenZZ
Feature activation+0.037
Top resid features:
L
TokenL
Feature activation-0.042
Top resid features:
IES
TokenIES
Feature activation+2.199
Top resid features:
_
Token_
Feature activation+0.000
Top resid features:
FB
TokenFB
Feature activation+0.000
Top resid features:
)
Token)
Feature activation+0.000
Top resid features:
became
Token became
Feature activation+0.000
Top resid features:
home
Token home
Feature activation+0.000
Top resid features:
itz
Tokenitz
Feature activation+0.154
Top resid features:
man
Tokenman
Feature activation+0.018
Top resid features:
,
Token,
Feature activation-0.089
Top resid features:
the
Token the
Feature activation-0.096
Top resid features:
W
Token W
Feature activation+0.052
Top resid features:
ies
Tokenies
Feature activation+2.617
Top resid features:
enthal
Tokenenthal
Feature activation+0.000
Top resid features:
Center
Token Center
Feature activation+0.000
Top resid features:
âĢ
TokenâĢ
Feature activation+0.000
Top resid features:
Ļ
TokenĻ
Feature activation+0.000
Top resid features:
s
Tokens
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+1.064
Top resid features:
featured
Token featured
Feature activation-0.043
Top resid features:
two
Token two
Feature activation+0.070
Top resid features:
stunning
Token stunning
Feature activation+0.054
Top resid features:
ups
Token ups
Feature activation+0.401
Top resid features:
ets
Tokenets
Feature activation+1.854
Top resid features:
that
Token that
Feature activation+0.000
Top resid features:
bolstered
Token bolstered
Feature activation+0.000
Top resid features:
En
Token En
Feature activation+0.000
Top resid features:
V
TokenV
Feature activation+0.000
Top resid features:
y
Tokeny
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+1.344
Top resid features:
born
Token born
Feature activation+0.133
Top resid features:
in
Token in
Feature activation+0.083
Top resid features:
L
Token L
Feature activation-0.056
Top resid features:
le
Tokenle
Feature activation-0.008
Top resid features:
ida
Tokenida
Feature activation+1.858
Top resid features:
,
Token,
Feature activation+0.000
Top resid features:
Spain
Token Spain
Feature activation+0.000
Top resid features:
,
Token,
Feature activation+0.000
Top resid features:
the
Token the
Feature activation+0.000
Top resid features:
son
Token son
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+1.172
Top resid features:
clearly
Token clearly
Feature activation-0.117
Top resid features:
becoming
Token becoming
Feature activation+0.105
Top resid features:
more
Token more
Feature activation+0.033
Top resid features:
western
Token western
Feature activation+0.095
Top resid features:
ized
Tokenized
Feature activation+2.031
Top resid features:
at
Token at
Feature activation+0.000
Top resid features:
a
Token a
Feature activation+0.000
Top resid features:
rapid
Token rapid
Feature activation+0.000
Top resid features:
pace
Token pace
Feature activation+0.000
Top resid features:
.
Token.
Feature activation+0.000
Top resid features:
http
Token http
Feature activation-0.001
Top resid features:
://
Token://
Feature activation+0.143
Top resid features:
www
Tokenwww
Feature activation+0.148
Top resid features:
.
Token.
Feature activation-0.044
Top resid features:
em
Tokenem
Feature activation-0.051
Top resid features:
ily
Tokenily
Feature activation+2.441
Top resid features:
review
Tokenreview
Feature activation+0.000
Top resid features:
s
Tokens
Feature activation+0.000
Top resid features:
.
Token.
Feature activation+0.000
Top resid features:
com
Tokencom
Feature activation+0.000
Top resid features:
/
Token/
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.976
Top resid features:
especially
Token especially
Feature activation-0.044
Top resid features:
since
Token since
Feature activation+0.147
Top resid features:
Sergio
Token Sergio
Feature activation-0.009
Top resid features:
March
Token March
Feature activation+0.040
Top resid features:
ion
Tokenion
Feature activation+2.164
Top resid features:
ne
Tokenne
Feature activation+0.000
Top resid features:
has
Token has
Feature activation+0.000
Top resid features:
stated
Token stated
Feature activation+0.000
Top resid features:
that
Token that
Feature activation+0.000
Top resid features:
it
Token it
Feature activation+0.000
Top resid features:
LI
TokenLI
Feature activation-0.023
Top resid features:
Los
Token Los
Feature activation+0.075
Top resid features:
Angeles
Token Angeles
Feature activation+0.131
Top resid features:
(@
Token (@
Feature activation+0.138
Top resid features:
ul
Tokenul
Feature activation+0.134
Top resid features:
il
Tokenil
Feature activation+1.767
Top resid features:
os
Tokenos
Feature activation+0.000
Top resid features:
angel
Tokenangel
Feature activation+0.000
Top resid features:
es
Tokenes
Feature activation+0.000
Top resid features:
)
Token)
Feature activation+0.000
Top resid features:
January
Token January
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+1.135
Top resid features:
based
Token based
Feature activation+0.056
Top resid features:
on
Token on
Feature activation+0.121
Top resid features:
Us
Token Us
Feature activation+0.024
Top resid features:
en
Tokenen
Feature activation+0.241
Top resid features:
et
Tokenet
Feature activation+1.669
Top resid features:
that
Token that
Feature activation+0.000
Top resid features:
started
Token started
Feature activation+0.000
Top resid features:
in
Token in
Feature activation+0.000
Top resid features:
1995
Token 1995
Feature activation+0.000
Top resid features:
.
Token.
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+1.061
Top resid features:
ier
Tokenier
Feature activation+0.831
Top resid features:
Come
Token Come
Feature activation+0.111
Top resid features:
to
Token to
Feature activation-0.053
Top resid features:
Boulevard
Token Boulevard
Feature activation+0.010
Top resid features:
ier
Tokenier
Feature activation+1.277
Top resid features:
to
Token to
Feature activation+0.000
Top resid features:
get
Token get
Feature activation+0.000
Top resid features:
in
Token in
Feature activation+0.000
Top resid features:
touch
Token touch
Feature activation+0.000
Top resid features:
with
Token with
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+1.512
Top resid features:
Civil
Token Civil
Feature activation-0.014
Top resid features:
Aviation
Token Aviation
Feature activation-0.057
Top resid features:
Organization
Token Organization
Feature activation+0.126
Top resid features:
(
Token (
Feature activation+0.087
Top resid features:
ICA
TokenICA
Feature activation+1.568
Top resid features:
O
TokenO
Feature activation+0.000
Top resid features:
)
Token)
Feature activation+0.000
Top resid features:
audit
Token audit
Feature activation+0.000
Top resid features:
that
Token that
Feature activation+0.000
Top resid features:
it
Token it
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.884
Top resid features:
âĢ
TokenâĢ
Feature activation+0.076
Top resid features:
Ŀ
TokenĿ
Feature activation+0.193
Top resid features:
J
Token J
Feature activation-0.042
Top resid features:
at
Tokenat
Feature activation+0.102
Top resid features:
ia
Tokenia
Feature activation+2.007
Top resid features:
said
Token said
Feature activation+0.000
Top resid features:
.
Token.
Feature activation+0.000
Top resid features:
Ċ
TokenĊ
Feature activation+0.000
Top resid features:
Ċ
TokenĊ
Feature activation+0.000
Top resid features:
In
TokenIn
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+1.009
Top resid features:
football
Token football
Feature activation-0.100
Top resid features:
legend
Token legend
Feature activation+0.082
Top resid features:
Th
Token Th
Feature activation-0.112
Top resid features:
ier
Tokenier
Feature activation+0.135
Top resid features:
ry
Tokenry
Feature activation+2.185
Top resid features:
Henry
Token Henry
Feature activation+0.000
Top resid features:
,
Token,
Feature activation+0.000
Top resid features:
C
Token C
Feature activation+0.000
Top resid features:
ound
Tokenound
Feature activation+0.000
Top resid features:
oul
Tokenoul
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+1.042
Top resid features:
ai
Tokenai
Feature activation+0.194
Top resid features:
yan
Tokenyan
Feature activation+0.128
Top resid features:
super
Token super
Feature activation-0.124
Top resid features:
-
Token-
Feature activation+0.037
Top resid features:
ty
Tokenty
Feature activation+1.917
Top resid features:
ph
Tokenph
Feature activation+0.000
Top resid features:
oon
Tokenoon
Feature activation+0.000
Top resid features:
.
Token.
Feature activation+0.000
Top resid features:
I
Token I
Feature activation+0.000
Top resid features:
'm
Token'm
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+1.005
Top resid features:
remarks
Token remarks
Feature activation+0.049
Top resid features:
,
Token,
Feature activation+0.028
Top resid features:
Mr
Token Mr
Feature activation-0.029
Top resid features:
Mans
Token Mans
Feature activation+0.084
Top resid features:
our
Tokenour
Feature activation+2.052
Top resid features:
said
Token said
Feature activation+0.000
Top resid features:
he
Token he
Feature activation+0.000
Top resid features:
had
Token had
Feature activation+0.000
Top resid features:
secured
Token secured
Feature activation+0.000
Top resid features:
the
Token the
Feature activation+0.000
Top resid features:
,
Token,
Feature activation-0.086
Top resid features:
âĢ
TokenâĢ
Feature activation+0.108
Top resid features:
Ŀ
TokenĿ
Feature activation+0.043
Top resid features:
Men
Token Men
Feature activation-0.039
Top resid features:
z
Tokenz
Feature activation+0.009
Top resid features:
ies
Tokenies
Feature activation+2.588
Top resid features:
said
Token said
Feature activation+0.000
Top resid features:
.
Token.
Feature activation+0.000
Top resid features:
âĢ
Token âĢ
Feature activation+0.000
Top resid features:
ľ
Tokenľ
Feature activation+0.000
Top resid features:
Whether
TokenWhether
Feature activation+0.000
Top resid features:

Decoder Weights Distribution

Head 0: 0.05

Head 1: 0.09

Head 2: 0.06

Head 3: 0.09

Head 4: 0.05

Head 5: 0.04

Head 6: 0.10

Head 7: 0.09

Head 8: 0.06

Head 9: 0.14

Head 10: 0.07

Head 11: 0.14

Positive logits

terday3.69

VIDIA3.36

confir3.36

proble3.22

lihood3.06

NetMessage3.02

condem2.83

destro2.82

ibilities2.66

etheless2.65

untary2.60

aying2.53

ktop2.50

seiz2.49

oreal2.46

andem2.46

nesday2.45

eatures2.45

hess2.42

orry2.42

Negative logits

Discussion-2.22

FactoryReloaded-2.18

arsity-2.17

izoph-2.07

Tokens-2.04

CLSID-1.98

�士-1.98

[|-1.90

Pers-1.83

shapeshifter-1.82

Jew-1.81

Hispanic-1.80

thous-1.80

Introduced-1.80

��-1.78

Nation-1.76

}.-1.75

foundland-1.75

----------1.73

BaseType-1.71

INTERVAL 1.758 - 1.953
CONTAINS 0.001%

INTERVAL 1.562 - 1.758
CONTAINS 0.003%

INTERVAL 1.367 - 1.562
CONTAINS 0.005%

INTERVAL 1.172 - 1.367
CONTAINS 0.007%

INTERVAL 0.977 - 1.172
CONTAINS 0.007%

INTERVAL 0.781 - 0.977
CONTAINS 0.010%

INTERVAL 0.586 - 0.781
CONTAINS 0.015%

<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
LI
TokenLI
Feature activation+0.008
Los
Token Los
Feature activation+0.000
Angeles
Token Angeles
Feature activation+0.000
(@
Token (@
Feature activation+0.000
ul
Tokenul
Feature activation+0.686
il
Tokenil
Feature activation+0.376
os
Tokenos
Feature activation+0.129
angel
Tokenangel
Feature activation+0.000
es
Tokenes
Feature activation+0.000
)
Token)
Feature activation+0.000

INTERVAL 0.391 - 0.586
CONTAINS 0.016%

<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
born
Token born
Feature activation+0.000
in
Token in
Feature activation+0.000
L
Token L
Feature activation+0.000
le
Tokenle
Feature activation+0.000
ida
Tokenida
Feature activation+0.463
,
Token,
Feature activation+0.000
Spain
Token Spain
Feature activation+0.000
,
Token,
Feature activation+0.000
the
Token the
Feature activation+0.000
son
Token son
Feature activation+0.000
with
Token with
Feature activation+0.000
walk
Token walk
Feature activation+0.000
ie
Tokenie
Feature activation+0.833
-
Token-
Feature activation+0.000
talk
Tokentalk
Feature activation+0.000
ies
Tokenies
Feature activation+0.524
to
Token to
Feature activation+0.000
update
Token update
Feature activation+0.000
the
Token the
Feature activation+0.000
crew
Token crew
Feature activation+0.000
about
Token about
Feature activation+0.000
http
Token http
Feature activation+0.000
://
Token://
Feature activation+0.000
www
Tokenwww
Feature activation+0.000
.
Token.
Feature activation+0.000
em
Tokenem
Feature activation+0.000
ily
Tokenily
Feature activation+0.405
review
Tokenreview
Feature activation+0.000
s
Tokens
Feature activation+0.000
.
Token.
Feature activation+0.000
com
Tokencom
Feature activation+0.000
/
Token/
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
featured
Token featured
Feature activation+0.000
two
Token two
Feature activation+0.000
stunning
Token stunning
Feature activation+0.000
ups
Token ups
Feature activation+0.000
ets
Tokenets
Feature activation+0.509
that
Token that
Feature activation+0.000
bolstered
Token bolstered
Feature activation+0.000
En
Token En
Feature activation+0.000
V
TokenV
Feature activation+0.000
y
Tokeny
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
G
TokenG
Feature activation+0.000
RI
TokenRI
Feature activation+0.000
ZZ
TokenZZ
Feature activation+0.000
L
TokenL
Feature activation+0.000
IES
TokenIES
Feature activation+0.523
_
Token_
Feature activation+0.000
FB
TokenFB
Feature activation+0.000
)
Token)
Feature activation+0.000
became
Token became
Feature activation+0.000
home
Token home
Feature activation+0.000

INTERVAL 0.195 - 0.391
CONTAINS 0.024%

<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
newcomer
Token newcomer
Feature activation+0.000
Lincoln
Token Lincoln
Feature activation+0.000
Ch
Token Ch
Feature activation+0.000
af
Tokenaf
Feature activation+0.000
ee
Tokenee
Feature activation+0.201
make
Token make
Feature activation+0.000
a
Token a
Feature activation+0.000
dent
Token dent
Feature activation+0.000
?
Token?
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
ai
Tokenai
Feature activation+1.078
yan
Tokenyan
Feature activation+0.421
super
Token super
Feature activation+0.000
-
Token-
Feature activation+0.000
ty
Tokenty
Feature activation+0.303
ph
Tokenph
Feature activation+0.000
oon
Tokenoon
Feature activation+0.000
.
Token.
Feature activation+0.000
I
Token I
Feature activation+0.000
'm
Token'm
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
Far
Token Far
Feature activation+0.000
oes
Tokenoes
Feature activation+0.272
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
set
Token set
Feature activation+0.000
of
Token of
Feature activation+0.000
islands
Token islands
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
called
Token called
Feature activation+0.000
Ap
Token Ap
Feature activation+0.000
od
Tokenod
Feature activation+0.000
ant
Tokenant
Feature activation+0.134
hes
Tokenhes
Feature activation+0.262
cas
Token cas
Feature activation+0.000
ear
Tokenear
Feature activation+0.000
iae
Tokeniae
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
,
Token,
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ŀ
TokenĿ
Feature activation+0.000
Men
Token Men
Feature activation+0.000
z
Tokenz
Feature activation+0.000
ies
Tokenies
Feature activation+0.296
said
Token said
Feature activation+0.000
.
Token.
Feature activation+0.000
âĢ
Token âĢ
Feature activation+0.000
ľ
Tokenľ
Feature activation+0.000
Whether
TokenWhether
Feature activation+0.000

INTERVAL 0.000 - 0.195
CONTAINS 99.913%

advice
Token advice
Feature activation+0.000
,
Token,
Feature activation+0.000
and
Token and
Feature activation+0.000
hints
Token hints
Feature activation+0.000
at
Token at
Feature activation+0.000
her
Token her
Feature activation+0.000
return
Token return
Feature activation+0.000
.
Token.
Feature activation+0.000
The
Token The
Feature activation+0.000
flashback
Token flashback
Feature activation+0.000
scene
Token scene
Feature activation+0.000
Ŀ
TokenĿ
Feature activation+0.000
,
Token,
Feature activation+0.000
he
Token he
Feature activation+0.000
says
Token says
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
S
TokenS
Feature activation+0.000
add
Tokenadd
Feature activation+0.000
am
Tokenam
Feature activation+0.000
constructed
Token constructed
Feature activation+0.000
his
Token his
Feature activation+0.000
plays
Token plays
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Wh
TokenWh
Feature activation+0.000
arton
Tokenarton
Feature activation+0.000
said
Token said
Feature activation+0.000
in
Token in
Feature activation+0.000
an
Token an
Feature activation+0.000
email
Token email
Feature activation+0.000
decade
Token decade
Feature activation+0.000
.
Token.
Feature activation+0.000
I
Token I
Feature activation+0.000
saw
Token saw
Feature activation+0.000
little
Token little
Feature activation+0.000
to
Token to
Feature activation+0.000
celebrate
Token celebrate
Feature activation+0.000
,
Token,
Feature activation+0.000
despite
Token despite
Feature activation+0.000
the
Token the
Feature activation+0.000
historical
Token historical
Feature activation+0.000
captures
Token captures
Feature activation+0.000
a
Token a
Feature activation+0.000
wider
Token wider
Feature activation+0.000
angle
Token angle
Feature activation+0.000
of
Token of
Feature activation+0.000
view
Token view
Feature activation+0.000
,
Token,
Feature activation+0.000
requiring
Token requiring
Feature activation+0.000
the
Token the
Feature activation+0.000
viewer
Token viewer
Feature activation+0.000
to
Token to
Feature activation+0.000

BOTTOM ACTIVATIONS
MIN = 0.000

regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
anking
Tokenanking
Feature activation+0.000
the
Token the
Feature activation+0.000
sc
Token sc
Feature activation+0.000
iss
Tokeniss
Feature activation+0.000
ile
Tokenile
Feature activation+0.000
bonds
Token bonds
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
anking
Tokenanking
Feature activation+0.000
the
Token the
Feature activation+0.000
sc
Token sc
Feature activation+0.000
iss
Tokeniss
Feature activation+0.000
ile
Tokenile
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
anking
Tokenanking
Feature activation+0.000
the
Token the
Feature activation+0.000
sc
Token sc
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
anking
Tokenanking
Feature activation+0.000
the
Token the
Feature activation+0.000
sc
Token sc
Feature activation+0.000
iss
Tokeniss
Feature activation+0.000
ile
Tokenile
Feature activation+1.451
(
Token (
Feature activation+0.000
Michael
TokenMichael
Feature activation+0.000
C
Token C
Feature activation+0.000
aine
Tokenaine
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
ile
Tokenile
Feature activation+1.451
(
Token (
Feature activation+0.000
Michael
TokenMichael
Feature activation+0.000
C
Token C
Feature activation+0.000
aine
Tokenaine
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
(
Token (
Feature activation+0.000
Michael
TokenMichael
Feature activation+0.000
C
Token C
Feature activation+0.000
aine
Tokenaine
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Michael
TokenMichael
Feature activation+0.000
C
Token C
Feature activation+0.000
aine
Tokenaine
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
anking
Tokenanking
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
anking
Tokenanking
Feature activation+0.000
the
Token the
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
C
Token C
Feature activation+0.000
aine
Tokenaine
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
aine
Tokenaine
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000

Top feature 3 in H0.11: (feature 12137

TOP ACTIVATIONS
MAX = 1.659

<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
iy
Tokeniy
Feature activation+0.000
ot
Tokenot
Feature activation+0.472
B
Token B
Feature activation+0.000
ons
Tokenons
Feature activation+1.383
ai
Tokenai
Feature activation+1.659
âĢĵ
Token âĢĵ
Feature activation+0.000
26
Token 26
Feature activation+0.000
Ben
Token Ben
Feature activation+0.000
Sir
Token Sir
Feature activation+0.000
a
Tokena
Feature activation+0.000
G
TokenG
Feature activation+0.000
ed
Tokened
Feature activation+0.096
im
Tokenim
Feature activation+1.186
inas
Tokeninas
Feature activation+0.988
J
Token J
Feature activation+0.000
urg
Tokenurg
Feature activation+1.594
ait
Tokenait
Feature activation+1.065
is
Tokenis
Feature activation+0.846
-
Token -
Feature activation+0.000
bass
Token bass
Feature activation+0.000
guitar
Token guitar
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
og
Tokenog
Feature activation+0.213
and
Token and
Feature activation+0.000
egg
Token egg
Feature activation+0.000
n
Tokenn
Feature activation+0.544
og
Tokenog
Feature activation+1.486
and
Token and
Feature activation+0.000
so
Token so
Feature activation+0.000
on
Token on
Feature activation+0.000
.
Token.
Feature activation+0.000
Any
Token Any
Feature activation+0.000
k
Tokenk
Feature activation+0.000
á
Tokená
Feature activation+0.662
H
Token H
Feature activation+0.000
ru
Tokenru
Feature activation+1.605
Å¡
Tokenš
Feature activation+0.340
ka
Tokenka
Feature activation+1.484
?
Token?
Feature activation+0.000
Yes
Token Yes
Feature activation+0.000
please
Token please
Feature activation+0.000
!
Token!
Feature activation+0.000
twitter
Token twitter
Feature activation+0.000
Franc
TokenFranc
Feature activation+0.000
is
Tokenis
Feature activation+0.066
N
Token N
Feature activation+0.000
gan
Tokengan
Feature activation+1.079
n
Tokenn
Feature activation+0.710
ou
Tokenou
Feature activation+1.439
($
Token ($
Feature activation+0.000
10
Token10
Feature activation+0.000
,
Token,
Feature activation+0.000
000
Token000
Feature activation+0.000
+
Token +
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
Teen
Token Teen
Feature activation+0.000
Che
Token Che
Feature activation+0.000
ez
Tokenez
Feature activation+0.516
Kab
Token Kab
Feature activation+0.108
hi
Tokenhi
Feature activation+1.428
Und
Token Und
Feature activation+0.000
erest
Tokenerest
Feature activation+0.186
imate
Tokenimate
Feature activation+0.292
Mat
Token Mat
Feature activation+0.000
K
Token K
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
Ar
Token Ar
Feature activation+0.000
be
Tokenbe
Feature activation+0.517
its
Tokenits
Feature activation+1.258
p
Tokenp
Feature activation+0.737
ap
Tokenap
Feature activation+1.411
ier
Tokenier
Feature activation+1.336
z
Token z
Feature activation+0.007
u
Tokenu
Feature activation+0.818
ver
Token ver
Feature activation+0.000
l
Tokenl
Feature activation+0.331
.,
Token.,
Feature activation+0.229
B
Token B
Feature activation+0.000
ial
Tokenial
Feature activation+1.155
y
Tokeny
Feature activation+0.420
st
Tokenst
Feature activation+0.744
ok
Tokenok
Feature activation+1.405
,
Token,
Feature activation+0.000
2006
Token 2006
Feature activation+0.000
;
Token;
Feature activation+0.000
D
Token D
Feature activation+0.000
ye
Tokenye
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
AN
Token AN
Feature activation+0.000
OTHER
TokenOTHER
Feature activation+0.509
R
Token R
Feature activation+0.000
OUND
TokenOUND
Feature activation+1.003
OF
TokenOF
Feature activation+1.401
SH
Token SH
Feature activation+0.027
IVER
TokenIVER
Feature activation+0.782
ING
TokenING
Feature activation+0.389
AND
Token AND
Feature activation+0.000
S
Token S
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
k
Token k
Feature activation+0.000
ms
Tokenms
Feature activation+0.505
from
Token from
Feature activation+0.000
In
Token In
Feature activation+0.000
uk
Tokenuk
Feature activation+1.394
ju
Tokenju
Feature activation+0.697
ak
Tokenak
Feature activation+0.951
,
Token,
Feature activation+0.000
Que
Token Que
Feature activation+0.000
.
Token.
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
,
Token,
Feature activation+0.000
like
Token like
Feature activation+0.000
chlor
Token chlor
Feature activation+0.000
p
Tokenp
Feature activation+0.324
yr
Tokenyr
Feature activation+1.368
if
Tokenif
Feature activation+1.179
os
Tokenos
Feature activation+0.950
and
Token and
Feature activation+0.000
carb
Token carb
Feature activation+0.000
of
Tokenof
Feature activation+0.623
Ar
Token Ar
Feature activation+0.000
be
Tokenbe
Feature activation+0.517
its
Tokenits
Feature activation+1.258
p
Tokenp
Feature activation+0.737
ap
Tokenap
Feature activation+1.411
ier
Tokenier
Feature activation+1.336
z
Token z
Feature activation+0.007
u
Tokenu
Feature activation+0.818
ver
Token ver
Feature activation+0.000
l
Tokenl
Feature activation+0.331
ä
Tokenä
Feature activation+0.696
Me
Token Me
Feature activation+0.000
iss
Tokeniss
Feature activation+1.156
en
Tokenen
Feature activation+0.660
-
Token -
Feature activation+0.000
G
Token G
Feature activation+0.000
ave
Tokenave
Feature activation+1.334
Ult
Token Ult
Feature activation+0.000
im
Tokenim
Feature activation+0.543
ogen
Tokenogen
Feature activation+0.554
iture
Tokeniture
Feature activation+0.214
to
Token to
Feature activation+0.000
Ap
Token Ap
Feature activation+0.000
od
Tokenod
Feature activation+1.021
ant
Tokenant
Feature activation+1.062
hes
Tokenhes
Feature activation+0.983
cas
Token cas
Feature activation+0.000
ear
Tokenear
Feature activation+1.323
iae
Tokeniae
Feature activation+0.145
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
An
Token An
Feature activation+0.000
ant
Tokenant
Feature activation+0.898
ap
Tokenap
Feature activation+1.120
ur
Tokenur
Feature activation+1.060
V
Token V
Feature activation+0.000
ign
Tokenign
Feature activation+1.320
an
Tokenan
Feature activation+0.876
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
s
Tokens
Feature activation+0.000
Foundation
Token Foundation
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
Ŀ
TokenĿ
Feature activation+0.000
âĢ
Token âĢ
Feature activation+0.000
ľ
Tokenľ
Feature activation+0.000
Car
TokenCar
Feature activation+0.458
ol
Tokenol
Feature activation+1.319
âĢ
TokenâĢ
Feature activation+0.000
Ŀ
TokenĿ
Feature activation+0.000
and
Token and
Feature activation+0.000
âĢ
Token âĢ
Feature activation+0.000
ľ
Tokenľ
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
âĢ
Token âĢ
Feature activation+0.000
ľ
Tokenľ
Feature activation+0.000
ston
Tokenston
Feature activation+1.030
et
Tokenet
Feature activation+0.853
ear
Tokenear
Feature activation+1.314
âĢ
TokenâĢ
Feature activation+0.000
Ŀ
TokenĿ
Feature activation+0.000
directly
Token directly
Feature activation+0.000
to
Token to
Feature activation+0.000
Comb
Token Comb
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
born
Token born
Feature activation+0.000
in
Token in
Feature activation+0.000
L
Token L
Feature activation+0.000
le
Tokenle
Feature activation+0.942
ida
Tokenida
Feature activation+1.309
,
Token,
Feature activation+0.000
Spain
Token Spain
Feature activation+0.000
,
Token,
Feature activation+0.000
the
Token the
Feature activation+0.000
son
Token son
Feature activation+0.000
U
TokenU
Feature activation+0.000
âĢĶ
Token âĢĶ
Feature activation+0.000
Ale
Token Ale
Feature activation+0.000
f
Tokenf
Feature activation+0.443
L
Token L
Feature activation+0.000
ila
Tokenila
Feature activation+1.302
Wa
Token Wa
Feature activation+0.000
L
Token L
Feature activation+0.000
ila
Tokenila
Feature activation+1.219
(@
Token (@
Feature activation+0.000
_
Token_
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
ent
Tokenent
Feature activation+0.000
i
Tokeni
Feature activation+0.000
pub
Token pub
Feature activation+0.000
b
Tokenb
Feature activation+0.214
lic
Tokenlic
Feature activation+1.298
i
Tokeni
Feature activation+0.435
inf
Token inf
Feature activation+0.000
ed
Tokened
Feature activation+0.286
eli
Tokeneli
Feature activation+0.083
.
Token.
Feature activation+0.000

Top DFA by src position
MAX = 0.514

<|endoftext|>
Token<|endoftext|>
Feature activation+0.252
Top resid features:
iy
Tokeniy
Feature activation+0.112
Top resid features:
ot
Tokenot
Feature activation+0.179
Top resid features:
B
Token B
Feature activation+0.514
Top resid features:
ons
Tokenons
Feature activation+0.079
Top resid features:
ai
Tokenai
Feature activation+0.147
Top resid features:
âĢĵ
Token âĢĵ
Feature activation+0.000
Top resid features:
26
Token 26
Feature activation+0.000
Top resid features:
Ben
Token Ben
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.128
Top resid features:
G
TokenG
Feature activation+0.358
Top resid features:
ed
Tokened
Feature activation+0.124
Top resid features:
im
Tokenim
Feature activation+0.177
Top resid features:
inas
Tokeninas
Feature activation-0.006
Top resid features:
J
Token J
Feature activation+0.168
Top resid features:
urg
Tokenurg
Feature activation+0.269
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.188
Top resid features:
og
Tokenog
Feature activation+0.033
Top resid features:
and
Token and
Feature activation+0.172
Top resid features:
egg
Token egg
Feature activation+0.261
Top resid features:
n
Tokenn
Feature activation+0.226
Top resid features:
og
Tokenog
Feature activation+0.231
Top resid features:
and
Token and
Feature activation+0.000
Top resid features:
so
Token so
Feature activation+0.000
Top resid features:
on
Token on
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.241
Top resid features:
k
Tokenk
Feature activation+0.318
Top resid features:
á
Tokená
Feature activation+0.109
Top resid features:
H
Token H
Feature activation+0.246
Top resid features:
ru
Tokenru
Feature activation+0.092
Top resid features:
Å¡
Tokenš
Feature activation-0.045
Top resid features:
ka
Tokenka
Feature activation+0.148
Top resid features:
Franc
TokenFranc
Feature activation+0.154
Top resid features:
is
Tokenis
Feature activation+0.145
Top resid features:
N
Token N
Feature activation+0.187
Top resid features:
gan
Tokengan
Feature activation+0.030
Top resid features:
n
Tokenn
Feature activation+0.114
Top resid features:
ou
Tokenou
Feature activation+0.219
Top resid features:
($
Token ($
Feature activation+0.000
Top resid features:
10
Token10
Feature activation+0.000
Top resid features:
,
Token,
Feature activation+0.000
Top resid features:
000
Token000
Feature activation+0.000
Top resid features:
+
Token +
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.276
Top resid features:
Teen
Token Teen
Feature activation+0.160
Top resid features:
Che
Token Che
Feature activation+0.159
Top resid features:
ez
Tokenez
Feature activation+0.092
Top resid features:
Kab
Token Kab
Feature activation+0.132
Top resid features:
hi
Tokenhi
Feature activation+0.234
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.169
Top resid features:
Ar
Token Ar
Feature activation+0.298
Top resid features:
be
Tokenbe
Feature activation+0.159
Top resid features:
its
Tokenits
Feature activation+0.128
Top resid features:
p
Tokenp
Feature activation+0.153
Top resid features:
ap
Tokenap
Feature activation+0.128
Top resid features:
ier
Tokenier
Feature activation+0.000
Top resid features:
.,
Token.,
Feature activation+0.017
Top resid features:
B
Token B
Feature activation+0.271
Top resid features:
ial
Tokenial
Feature activation-0.032
Top resid features:
y
Tokeny
Feature activation+0.013
Top resid features:
st
Tokenst
Feature activation+0.067
Top resid features:
ok
Tokenok
Feature activation+0.363
Top resid features:
,
Token,
Feature activation+0.000
Top resid features:
2006
Token 2006
Feature activation+0.000
Top resid features:
;
Token;
Feature activation+0.000
Top resid features:
D
Token D
Feature activation+0.000
Top resid features:
ye
Tokenye
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.181
Top resid features:
AN
Token AN
Feature activation+0.303
Top resid features:
OTHER
TokenOTHER
Feature activation-0.010
Top resid features:
R
Token R
Feature activation+0.277
Top resid features:
OUND
TokenOUND
Feature activation+0.066
Top resid features:
OF
TokenOF
Feature activation+0.208
Top resid features:
SH
Token SH
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.168
Top resid features:
k
Token k
Feature activation+0.169
Top resid features:
ms
Tokenms
Feature activation+0.117
Top resid features:
from
Token from
Feature activation+0.084
Top resid features:
In
Token In
Feature activation+0.159
Top resid features:
uk
Tokenuk
Feature activation+0.322
Top resid features:
ju
Tokenju
Feature activation+0.000
Top resid features:
ak
Tokenak
Feature activation+0.000
Top resid features:
,
Token,
Feature activation+0.000
Top resid features:
Que
Token Que
Feature activation+0.000
Top resid features:
.
Token.
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.083
Top resid features:
,
Token,
Feature activation+0.081
Top resid features:
like
Token like
Feature activation+0.132
Top resid features:
chlor
Token chlor
Feature activation+0.033
Top resid features:
p
Tokenp
Feature activation+0.239
Top resid features:
yr
Tokenyr
Feature activation+0.424
Top resid features:
if
Tokenif
Feature activation+0.000
Top resid features:
os
Tokenos
Feature activation+0.000
Top resid features:
and
Token and
Feature activation+0.000
Top resid features:
carb
Token carb
Feature activation+0.000
Top resid features:
of
Tokenof
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.239
Top resid features:
Ar
Token Ar
Feature activation+0.195
Top resid features:
be
Tokenbe
Feature activation+0.133
Top resid features:
its
Tokenits
Feature activation+0.108
Top resid features:
p
Tokenp
Feature activation+0.187
Top resid features:
ap
Tokenap
Feature activation+0.064
Top resid features:
to
Token to
Feature activation+0.027
Top resid features:
Me
Token Me
Feature activation+0.117
Top resid features:
iss
Tokeniss
Feature activation+0.017
Top resid features:
en
Tokenen
Feature activation+0.021
Top resid features:
-
Token -
Feature activation-0.089
Top resid features:
G
Token G
Feature activation+0.337
Top resid features:
ave
Tokenave
Feature activation+0.193
Top resid features:
Ult
Token Ult
Feature activation+0.000
Top resid features:
im
Tokenim
Feature activation+0.000
Top resid features:
ogen
Tokenogen
Feature activation+0.000
Top resid features:
iture
Tokeniture
Feature activation+0.000
Top resid features:
Ap
Token Ap
Feature activation+0.090
Top resid features:
od
Tokenod
Feature activation+0.173
Top resid features:
ant
Tokenant
Feature activation+0.004
Top resid features:
hes
Tokenhes
Feature activation+0.029
Top resid features:
cas
Token cas
Feature activation+0.027
Top resid features:
ear
Tokenear
Feature activation+0.332
Top resid features:
iae
Tokeniae
Feature activation+0.000
Top resid features:
.
Token.
Feature activation+0.000
Top resid features:
Ċ
TokenĊ
Feature activation+0.000
Top resid features:
Ċ
TokenĊ
Feature activation+0.000
Top resid features:
The
TokenThe
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.210
Top resid features:
,
Token,
Feature activation+0.084
Top resid features:
An
Token An
Feature activation+0.167
Top resid features:
ant
Tokenant
Feature activation+0.058
Top resid features:
ap
Tokenap
Feature activation+0.088
Top resid features:
ur
Tokenur
Feature activation+0.078
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.096
Top resid features:
Ŀ
TokenĿ
Feature activation+0.173
Top resid features:
âĢ
Token âĢ
Feature activation+0.026
Top resid features:
ľ
Tokenľ
Feature activation+0.132
Top resid features:
Car
TokenCar
Feature activation+0.235
Top resid features:
ol
Tokenol
Feature activation+0.282
Top resid features:
âĢ
TokenâĢ
Feature activation+0.000
Top resid features:
Ŀ
TokenĿ
Feature activation+0.000
Top resid features:
and
Token and
Feature activation+0.000
Top resid features:
âĢ
Token âĢ
Feature activation+0.000
Top resid features:
ľ
Tokenľ
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.121
Top resid features:
âĢ
Token âĢ
Feature activation+0.104
Top resid features:
ľ
Tokenľ
Feature activation+0.190
Top resid features:
ston
Tokenston
Feature activation+0.125
Top resid features:
et
Tokenet
Feature activation+0.087
Top resid features:
ear
Tokenear
Feature activation+0.311
Top resid features:
âĢ
TokenâĢ
Feature activation+0.000
Top resid features:
Ŀ
TokenĿ
Feature activation+0.000
Top resid features:
directly
Token directly
Feature activation+0.000
Top resid features:
to
Token to
Feature activation+0.000
Top resid features:
Comb
Token Comb
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.057
Top resid features:
born
Token born
Feature activation+0.163
Top resid features:
in
Token in
Feature activation+0.148
Top resid features:
L
Token L
Feature activation+0.308
Top resid features:
le
Tokenle
Feature activation+0.071
Top resid features:
ida
Tokenida
Feature activation+0.185
Top resid features:
,
Token,
Feature activation+0.000
Top resid features:
Spain
Token Spain
Feature activation+0.000
Top resid features:
,
Token,
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.112
Top resid features:
U
TokenU
Feature activation+0.184
Top resid features:
âĢĶ
Token âĢĶ
Feature activation+0.063
Top resid features:
Ale
Token Ale
Feature activation+0.042
Top resid features:
f
Tokenf
Feature activation+0.043
Top resid features:
L
Token L
Feature activation+0.255
Top resid features:
ila
Tokenila
Feature activation+0.229
Top resid features:
Wa
Token Wa
Feature activation+0.000
Top resid features:
L
Token L
Feature activation+0.000
Top resid features:
ila
Tokenila
Feature activation+0.000
Top resid features:
(@
Token (@
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.154
Top resid features:
ent
Tokenent
Feature activation+0.225
Top resid features:
i
Tokeni
Feature activation+0.096
Top resid features:
pub
Token pub
Feature activation+0.109
Top resid features:
b
Tokenb
Feature activation+0.144
Top resid features:
lic
Tokenlic
Feature activation+0.193
Top resid features:
i
Tokeni
Feature activation+0.000
Top resid features:

Decoder Weights Distribution

Head 0: 0.05

Head 1: 0.13

Head 2: 0.07

Head 3: 0.13

Head 4: 0.04

Head 5: 0.04

Head 6: 0.06

Head 7: 0.05

Head 8: 0.06

Head 9: 0.16

Head 10: 0.07

Head 11: 0.14

Positive logits

NetMessage3.28

terness2.88

terday2.77

lihood2.64

proble2.61

bda2.54

hess2.54

etheless2.53

esson2.52

SourceFile2.49

odies2.44

oint2.42

verend2.42

pmwiki2.41

confir2.40

2.38

hes2.37

ands2.37

orry2.36

LOS2.36

Negative logits

CLSID-2.75

foundland-2.75

Discussion-2.30

Hilbert-2.27

arsity-2.26

SetTextColor-2.25

[|-2.24

BaseType-2.15

Period-2.09

referen-2.08

ItemLevel-2.05

thous-2.02

continuum-1.97

Fiesta-1.96

Tokens-1.95

Came-1.94

Daytona-1.94

Track-1.93

paired-1.90

bracelet-1.90

INTERVAL 1.493 - 1.659
CONTAINS 0.001%

G
TokenG
Feature activation+0.000
ed
Tokened
Feature activation+0.096
im
Tokenim
Feature activation+1.186
inas
Tokeninas
Feature activation+0.988
J
Token J
Feature activation+0.000
urg
Tokenurg
Feature activation+1.594
ait
Tokenait
Feature activation+1.065
is
Tokenis
Feature activation+0.846
-
Token -
Feature activation+0.000
bass
Token bass
Feature activation+0.000
guitar
Token guitar
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
iy
Tokeniy
Feature activation+0.000
ot
Tokenot
Feature activation+0.472
B
Token B
Feature activation+0.000
ons
Tokenons
Feature activation+1.383
ai
Tokenai
Feature activation+1.659
âĢĵ
Token âĢĵ
Feature activation+0.000
26
Token 26
Feature activation+0.000
Ben
Token Ben
Feature activation+0.000
Sir
Token Sir
Feature activation+0.000
a
Tokena
Feature activation+0.000

INTERVAL 1.327 - 1.493
CONTAINS 0.003%

<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
Ar
Token Ar
Feature activation+0.000
be
Tokenbe
Feature activation+0.517
its
Tokenits
Feature activation+1.258
p
Tokenp
Feature activation+0.737
ap
Tokenap
Feature activation+1.411
ier
Tokenier
Feature activation+1.336
z
Token z
Feature activation+0.007
u
Tokenu
Feature activation+0.818
ver
Token ver
Feature activation+0.000
l
Tokenl
Feature activation+0.331
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
k
Token k
Feature activation+0.000
ms
Tokenms
Feature activation+0.505
from
Token from
Feature activation+0.000
In
Token In
Feature activation+0.000
uk
Tokenuk
Feature activation+1.394
ju
Tokenju
Feature activation+0.697
ak
Tokenak
Feature activation+0.951
,
Token,
Feature activation+0.000
Que
Token Que
Feature activation+0.000
.
Token.
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
og
Tokenog
Feature activation+0.213
and
Token and
Feature activation+0.000
egg
Token egg
Feature activation+0.000
n
Tokenn
Feature activation+0.544
og
Tokenog
Feature activation+1.486
and
Token and
Feature activation+0.000
so
Token so
Feature activation+0.000
on
Token on
Feature activation+0.000
.
Token.
Feature activation+0.000
Any
Token Any
Feature activation+0.000
Franc
TokenFranc
Feature activation+0.000
is
Tokenis
Feature activation+0.066
N
Token N
Feature activation+0.000
gan
Tokengan
Feature activation+1.079
n
Tokenn
Feature activation+0.710
ou
Tokenou
Feature activation+1.439
($
Token ($
Feature activation+0.000
10
Token10
Feature activation+0.000
,
Token,
Feature activation+0.000
000
Token000
Feature activation+0.000
+
Token +
Feature activation+0.000
Ar
Token Ar
Feature activation+0.000
be
Tokenbe
Feature activation+0.517
its
Tokenits
Feature activation+1.258
p
Tokenp
Feature activation+0.737
ap
Tokenap
Feature activation+1.411
ier
Tokenier
Feature activation+1.336
z
Token z
Feature activation+0.007
u
Tokenu
Feature activation+0.818
ver
Token ver
Feature activation+0.000
l
Tokenl
Feature activation+0.331
ä
Tokenä
Feature activation+0.696

INTERVAL 1.161 - 1.327
CONTAINS 0.008%

<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
L
Token L
Feature activation+0.000
ian
Tokenian
Feature activation+0.857
y
Tokeny
Feature activation+0.589
ung
Tokenung
Feature activation+1.472
ang
Tokenang
Feature activation+1.264
,
Token,
Feature activation+0.000
east
Token east
Feature activation+0.000
China
Token China
Feature activation+0.000
's
Token's
Feature activation+0.000
Jiang
Token Jiang
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
iq
Tokeniq
Feature activation+0.000
ar
Tokenar
Feature activation+0.236
Ali
Token Ali
Feature activation+0.006
Bh
Token Bh
Feature activation+0.000
ut
Tokenut
Feature activation+1.283
to
Tokento
Feature activation+0.483
.
Token.
Feature activation+0.000
By
Token By
Feature activation+0.000
this
Token this
Feature activation+0.000
point
Token point
Feature activation+0.000
Form
Token Form
Feature activation+0.000
an
Tokenan
Feature activation+0.474
and
Token and
Feature activation+0.000
Michael
Token Michael
Feature activation+0.000
J
Token J
Feature activation+0.000
aven
Tokenaven
Feature activation+1.168
Fort
Token Fort
Feature activation+0.000
ner
Tokenner
Feature activation+0.717
,
Token,
Feature activation+0.000
author
Token author
Feature activation+0.000
of
Token of
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
the
Token the
Feature activation+0.000
coast
Token coast
Feature activation+0.000
at
Token at
Feature activation+0.000
H
Token H
Feature activation+0.000
umb
Tokenumb
Feature activation+1.203
old
Tokenold
Feature activation+0.956
t
Tokent
Feature activation+0.316
.
Token.
Feature activation+0.000
The
Token The
Feature activation+0.000
key
Token key
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
engine
Token engine
Feature activation+0.000
.
Token.
Feature activation+0.000
This
Token This
Feature activation+0.000
V
Token V
Feature activation+0.000
ost
Tokenost
Feature activation+1.166
ok
Tokenok
Feature activation+1.118
-
Token-
Feature activation+0.000
inspired
Tokeninspired
Feature activation+0.000
model
Token model
Feature activation+0.000
includes
Token includes
Feature activation+0.000

INTERVAL 0.995 - 1.161
CONTAINS 0.018%

<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
and
Token and
Feature activation+0.000
he
Token he
Feature activation+0.000
used
Token used
Feature activation+0.000
Ph
Token Ph
Feature activation+0.000
air
Tokenair
Feature activation+1.079
's
Token's
Feature activation+0.000
1993
Token 1993
Feature activation+0.000
debut
Token debut
Feature activation+0.000
Exile
Token Exile
Feature activation+0.000
in
Token in
Feature activation+0.000
Area
Token Area
Feature activation+0.000
resident
Token resident
Feature activation+0.000
Jeanne
Token Jeanne
Feature activation+0.000
Sol
Token Sol
Feature activation+0.000
n
Tokenn
Feature activation+0.502
ord
Tokenord
Feature activation+1.016
al
Tokenal
Feature activation+0.453
,
Token,
Feature activation+0.000
who
Token who
Feature activation+0.000
said
Token said
Feature activation+0.000
she
Token she
Feature activation+0.000
no
Token no
Feature activation+0.000
resistance
Token resistance
Feature activation+0.000
development
Token development
Feature activation+0.000
to
Token to
Feature activation+0.000
te
Token te
Feature activation+0.000
ix
Tokenix
Feature activation+0.996
ob
Tokenob
Feature activation+0.638
act
Tokenact
Feature activation+0.823
in
Tokenin
Feature activation+0.272
,"
Token,"
Feature activation+0.000
Lewis
Token Lewis
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
a
Token a
Feature activation+0.000
number
Token number
Feature activation+0.000
of
Token of
Feature activation+0.000
R
Token R
Feature activation+0.000
TS
TokenTS
Feature activation+1.083
games
Token games
Feature activation+0.000
,
Token,
Feature activation+0.000
and
Token and
Feature activation+0.000
like
Token like
Feature activation+0.000
most
Token most
Feature activation+0.000
-
Token-
Feature activation+0.000
field
Tokenfield
Feature activation+0.000
,
Token,
Feature activation+0.000
R
Token R
Feature activation+0.000
it
Tokenit
Feature activation+0.879
che
Tokenche
Feature activation+1.022
y
Tokeny
Feature activation+0.282
âĢĵ
TokenâĢĵ
Feature activation+0.000
Ch
TokenCh
Feature activation+0.000
r
Tokenr
Feature activation+0.000
ét
Tokenét
Feature activation+0.108

INTERVAL 0.829 - 0.995
CONTAINS 0.032%

<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
.
Token.
Feature activation+0.000
S
Token S
Feature activation+0.000
.
Token.
Feature activation+0.000
F
Token F
Feature activation+0.000
ORE
TokenORE
Feature activation+0.919
IGN
TokenIGN
Feature activation+0.469
POL
Token POL
Feature activation+0.000
IC
TokenIC
Feature activation+0.481
Y
TokenY
Feature activation+0.000
:
Token:
Feature activation+0.000
Sh
Token Sh
Feature activation+0.000
ola
Tokenola
Feature activation+1.213
aur
Token aur
Feature activation+0.000
Sh
Token Sh
Feature activation+0.000
ab
Tokenab
Feature activation+1.039
nam
Tokennam
Feature activation+0.969
and
Token and
Feature activation+0.000
A
Token A
Feature activation+0.000
ank
Tokenank
Feature activation+0.889
en
Tokenen
Feature activation+0.119
to
Token to
Feature activation+0.000
Sh
Token Sh
Feature activation+0.000
ab
Tokenab
Feature activation+1.039
nam
Tokennam
Feature activation+0.969
and
Token and
Feature activation+0.000
A
Token A
Feature activation+0.000
ank
Tokenank
Feature activation+0.889
en
Tokenen
Feature activation+0.119
to
Token to
Feature activation+0.000
his
Token his
Feature activation+0.000
name
Token name
Feature activation+0.000
said
Token said
Feature activation+0.000
a
Token a
Feature activation+0.000
student
Token student
Feature activation+0.000
of
Token of
Feature activation+0.000
Mark
Token Mark
Feature activation+0.000
F
Token F
Feature activation+0.000
on
Tokenon
Feature activation+0.936
stad
Tokenstad
Feature activation+0.743
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
Texas
Token Texas
Feature activation+0.000
State
Token State
Feature activation+0.000
The
Token The
Feature activation+0.000
manufacturer
Token manufacturer
Feature activation+0.000
Z
Token Z
Feature activation+0.000
d
Tokend
Feature activation+0.204
ur
Tokenur
Feature activation+0.940
ien
Tokenien
Feature activation+0.980
ci
Tokenci
Feature activation+0.618
k
Tokenk
Feature activation+0.141
has
Token has
Feature activation+0.000
a
Token a
Feature activation+0.000
bit
Token bit
Feature activation+0.000

INTERVAL 0.663 - 0.829
CONTAINS 0.045%

with
Token with
Feature activation+0.000
Greenwald
Token Greenwald
Feature activation+0.000
and
Token and
Feature activation+0.000
Po
Token Po
Feature activation+0.000
it
Tokenit
Feature activation+0.903
ras
Tokenras
Feature activation+0.793
for
Token for
Feature activation+0.000
a
Token a
Feature activation+0.000
number
Token number
Feature activation+0.000
of
Token of
Feature activation+0.000
recent
Token recent
Feature activation+0.000
the
Token the
Feature activation+0.000
hatred
Token hatred
Feature activation+0.000
of
Token of
Feature activation+0.000
E
Token E
Feature activation+0.000
w
Tokenw
Feature activation+0.355
oks
Tokenoks
Feature activation+0.817
stems
Token stems
Feature activation+0.000
from
Token from
Feature activation+0.000
that
Token that
Feature activation+0.000
unlikely
Token unlikely
Feature activation+0.000
-
Token-
Feature activation+0.000
M
Token M
Feature activation+0.000
FA
TokenFA
Feature activation+0.950
Armenia
Token Armenia
Feature activation+0.000
,
Token,
Feature activation+0.000
Sh
Token Sh
Feature activation+0.000
av
Tokenav
Feature activation+0.754
arsh
Tokenarsh
Feature activation+0.634
Koch
Token Koch
Feature activation+0.000
ary
Tokenary
Feature activation+0.444
an
Tokenan
Feature activation+0.016
said
Token said
Feature activation+0.000
-
Token-
Feature activation+0.000
fort
Tokenfort
Feature activation+0.013
ress
Tokenress
Feature activation+0.005
the
Token the
Feature activation+0.000
Bast
Token Bast
Feature activation+0.000
ille
Tokenille
Feature activation+0.746
,
Token,
Feature activation+0.000
seems
Token seems
Feature activation+0.000
entirely
Token entirely
Feature activation+0.000
appropriate
Token appropriate
Feature activation+0.000
to
Token to
Feature activation+0.000
goals
Token goals
Feature activation+0.000
scored
Token scored
Feature activation+0.000
by
Token by
Feature activation+0.000
H
Token H
Feature activation+0.000
rist
Tokenrist
Feature activation+1.158
ov
Tokenov
Feature activation+0.751
.
Token.
Feature activation+0.000
But
Token But
Feature activation+0.000
Saturday
Token Saturday
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000

INTERVAL 0.498 - 0.663
CONTAINS 0.052%

<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
spike
Token spike
Feature activation+0.000
"
Token"
Feature activation+0.000
that
Token that
Feature activation+0.000
v
Token v
Feature activation+0.000
ented
Tokenented
Feature activation+0.540
from
Token from
Feature activation+0.000
the
Token the
Feature activation+0.000
subs
Token subs
Feature activation+0.000
ur
Tokenur
Feature activation+0.228
face
Tokenface
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
,
Token,
Feature activation+0.000
as
Token as
Feature activation+0.000
did
Token did
Feature activation+0.000
Word
Token Word
Feature activation+0.327
worth
Tokenworth
Feature activation+0.501
.
Token.
Feature activation+0.000
Et
Token Et
Feature activation+0.000
c
Tokenc
Feature activation+0.000
eter
Tokeneter
Feature activation+0.554
a
Tokena
Feature activation+0.000
that
Token that
Feature activation+0.000
N
Token N
Feature activation+0.000
TT
TokenTT
Feature activation+0.895
Do
Token Do
Feature activation+0.000
Co
TokenCo
Feature activation+0.562
Mo
TokenMo
Feature activation+0.612
âĢ
Token âĢ
Feature activation+0.000
ľ
Tokenľ
Feature activation+0.000
is
Tokenis
Feature activation+0.463
expected
Token expected
Feature activation+0.000
to
Token to
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
side
Token side
Feature activation+0.000
and
Token and
Feature activation+0.000
"
Token "
Feature activation+0.000
SL
TokenSL
Feature activation+0.347
OW
TokenOW
Feature activation+0.547
"
Token"
Feature activation+0.000
on
Token on
Feature activation+0.000
the
Token the
Feature activation+0.000
other
Token other
Feature activation+0.000
.
Token.
Feature activation+0.000
mid
Token mid
Feature activation+0.000
-
Token-
Feature activation+0.000
March
TokenMarch
Feature activation+0.000
,
Token,
Feature activation+0.000
W
Token W
Feature activation+0.000
FP
TokenFP
Feature activation+0.620
has
Token has
Feature activation+0.000
provided
Token provided
Feature activation+0.000
life
Token life
Feature activation+0.000
-
Token-
Feature activation+0.000
saving
Tokensaving
Feature activation+0.000

INTERVAL 0.332 - 0.498
CONTAINS 0.064%

are
Token are
Feature activation+0.000
big
Token big
Feature activation+0.000
,
Token,
Feature activation+0.000
Steve
Token Steve
Feature activation+0.000
Ar
Token Ar
Feature activation+0.000
wood
Tokenwood
Feature activation+0.481
,
Token,
Feature activation+0.000
CEO
Token CEO
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
MED
Token MED
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
them
Token them
Feature activation+0.000
their
Token their
Feature activation+0.000
in
Token in
Feature activation+0.000
iqu
Tokeniqu
Feature activation+0.708
ities
Tokenities
Feature activation+0.451
,
Token,
Feature activation+0.000
and
Token and
Feature activation+0.000
admit
Token admit
Feature activation+0.000
them
Token them
Feature activation+0.000
into
Token into
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
Man
TokenMan
Feature activation+0.000
,
Token,
Feature activation+0.000
#
Token #
Feature activation+0.000
Black
TokenBlack
Feature activation+0.000
ish
Tokenish
Feature activation+0.482
looks
Token looks
Feature activation+0.000
so
Token so
Feature activation+0.000
original
Token original
Feature activation+0.000
!
Token!
Feature activation+0.000
A
Token A
Feature activation+0.000
to
Token to
Feature activation+0.000
complete
Token complete
Feature activation+0.000
âĢ
Token âĢ
Feature activation+0.000
ĺ
Tokenĺ
Feature activation+0.000
H
TokenH
Feature activation+0.000
ex
Tokenex
Feature activation+0.410
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
left
Token left
Feature activation+0.000
the
Token the
Feature activation+0.000
band
Token band
Feature activation+0.000
like
Token like
Feature activation+0.000
?
Token?
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ch
TokenCh
Feature activation+0.000
ops
Tokenops
Feature activation+0.386
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Allow
TokenAllow
Feature activation+0.000
me
Token me
Feature activation+0.000
to
Token to
Feature activation+0.000

INTERVAL 0.166 - 0.332
CONTAINS 0.080%

Sur
Token Sur
Feature activation+0.000
jit
Tokenjit
Feature activation+0.420
Singh
Token Singh
Feature activation+0.421
B
Token B
Feature activation+0.000
ades
Tokenades
Feature activation+0.800
ha
Tokenha
Feature activation+0.236
are
Token are
Feature activation+0.000
accused
Token accused
Feature activation+0.000
of
Token of
Feature activation+0.000
ordering
Token ordering
Feature activation+0.000
the
Token the
Feature activation+0.000
You
TokenYou
Feature activation+0.000
know
Token know
Feature activation+0.000
the
Token the
Feature activation+0.000
old
Token old
Feature activation+0.000
ad
Token ad
Feature activation+0.000
age
Tokenage
Feature activation+0.180
:
Token:
Feature activation+0.000
"
Token "
Feature activation+0.000
A
TokenA
Feature activation+0.000
picture
Token picture
Feature activation+0.000
's
Token's
Feature activation+0.000
of
Token of
Feature activation+0.000
F
Token F
Feature activation+0.000
amer
Tokenamer
Feature activation+0.239
Troy
Token Troy
Feature activation+0.000
A
Token A
Feature activation+0.000
ik
Tokenik
Feature activation+0.310
man
Tokenman
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
After
TokenAfter
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
And
TokenAnd
Feature activation+0.000
thr
Token thr
Feature activation+0.000
um
Tokenum
Feature activation+0.761
ming
Tokenming
Feature activation+1.221
underneath
Token underneath
Feature activation+0.193
the
Token the
Feature activation+0.000
outrage
Token outrage
Feature activation+0.000
about
Token about
Feature activation+0.000
this
Token this
Feature activation+0.000
interview
Token interview
Feature activation+0.000
Turns
Token Turns
Feature activation+0.000
13
Token 13
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Sp
TokenSp
Feature activation+0.000
aces
Tokenaces
Feature activation+0.245
M
Token M
Feature activation+0.000
oved
Tokenoved
Feature activation+0.244
6
Token 6
Feature activation+0.000
05
Token05
Feature activation+0.000
6
Token6
Feature activation+0.000

INTERVAL 0.000 - 0.166
CONTAINS 99.697%

higher
Token higher
Feature activation+0.000
risk
Token risk
Feature activation+0.000
of
Token of
Feature activation+0.000
developing
Token developing
Feature activation+0.000
diabetes
Token diabetes
Feature activation+0.000
,
Token,
Feature activation+0.000
they
Token they
Feature activation+0.000
should
Token should
Feature activation+0.000
pay
Token pay
Feature activation+0.000
more
Token more
Feature activation+0.000
attention
Token attention
Feature activation+0.000
need
Token need
Feature activation+0.000
to
Token to
Feature activation+0.000
do
Token do
Feature activation+0.000
,
Token,
Feature activation+0.000
it
Token it
Feature activation+0.000
could
Token could
Feature activation+0.000
be
Token be
Feature activation+0.000
hundreds
Token hundreds
Feature activation+0.000
or
Token or
Feature activation+0.000
thousands
Token thousands
Feature activation+0.000
of
Token of
Feature activation+0.000
Guardians
Token Guardians
Feature activation+0.000
coverage
Token coverage
Feature activation+0.000
here
Token here
Feature activation+0.000
at
Token at
Feature activation+0.000
Forbes
Token Forbes
Feature activation+0.000
Games
Token Games
Feature activation+0.000
,
Token,
Feature activation+0.000
I
Token I
Feature activation+0.000
loaded
Token loaded
Feature activation+0.000
up
Token up
Feature activation+0.000
my
Token my
Feature activation+0.000
addition
Token addition
Feature activation+0.000
to
Token to
Feature activation+0.000
a
Token a
Feature activation+0.000
traditional
Token traditional
Feature activation+0.000
bakery
Token bakery
Feature activation+0.000
and
Token and
Feature activation+0.000
brewery
Token brewery
Feature activation+0.000
,
Token,
Feature activation+0.000
the
Token the
Feature activation+0.000
space
Token space
Feature activation+0.000
is
Token is
Feature activation+0.000
through
Token through
Feature activation+0.000
which
Token which
Feature activation+0.000
an
Token an
Feature activation+0.000
individual
Token individual
Feature activation+0.000
relates
Token relates
Feature activation+0.000
to
Token to
Feature activation+0.000
,
Token,
Feature activation+0.000
perce
Token perce
Feature activation+0.000
ives
Tokenives
Feature activation+0.000
and
Token and
Feature activation+0.000
thinks
Token thinks
Feature activation+0.000

BOTTOM ACTIVATIONS
MIN = 0.000

by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
anking
Tokenanking
Feature activation+0.000
the
Token the
Feature activation+0.000
sc
Token sc
Feature activation+0.000
iss
Tokeniss
Feature activation+0.000
ile
Tokenile
Feature activation+0.000
bonds
Token bonds
Feature activation+0.000
.
Token.
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
anking
Tokenanking
Feature activation+0.000
the
Token the
Feature activation+0.000
sc
Token sc
Feature activation+0.000
iss
Tokeniss
Feature activation+0.000
ile
Tokenile
Feature activation+0.000
bonds
Token bonds
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
anking
Tokenanking
Feature activation+0.000
the
Token the
Feature activation+0.000
sc
Token sc
Feature activation+0.000
iss
Tokeniss
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
anking
Tokenanking
Feature activation+0.000
the
Token the
Feature activation+0.000
sc
Token sc
Feature activation+0.000
iss
Tokeniss
Feature activation+0.000
ile
Tokenile
Feature activation+0.000
(
Token (
Feature activation+0.000
Michael
TokenMichael
Feature activation+0.000
C
Token C
Feature activation+0.000
aine
Tokenaine
Feature activation+0.846
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
ile
Tokenile
Feature activation+0.007
(
Token (
Feature activation+0.000
Michael
TokenMichael
Feature activation+0.000
C
Token C
Feature activation+0.000
aine
Tokenaine
Feature activation+0.846
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Michael
TokenMichael
Feature activation+0.000
C
Token C
Feature activation+0.000
aine
Tokenaine
Feature activation+0.846
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
C
Token C
Feature activation+0.000
aine
Tokenaine
Feature activation+0.846
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
anking
Tokenanking
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
anking
Tokenanking
Feature activation+0.000
the
Token the
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
anking
Tokenanking
Feature activation+0.000
the
Token the
Feature activation+0.000
sc
Token sc
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
aine
Tokenaine
Feature activation+0.846
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000

Top feature 4 in H0.11: (feature 14135

TOP ACTIVATIONS
MAX = 0.000

regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
anking
Tokenanking
Feature activation+0.000
the
Token the
Feature activation+0.000
sc
Token sc
Feature activation+0.000
iss
Tokeniss
Feature activation+0.000
ile
Tokenile
Feature activation+0.000
bonds
Token bonds
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
anking
Tokenanking
Feature activation+0.000
the
Token the
Feature activation+0.000
sc
Token sc
Feature activation+0.000
iss
Tokeniss
Feature activation+0.000
ile
Tokenile
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
anking
Tokenanking
Feature activation+0.000
the
Token the
Feature activation+0.000
sc
Token sc
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
anking
Tokenanking
Feature activation+0.000
the
Token the
Feature activation+0.000
sc
Token sc
Feature activation+0.000
iss
Tokeniss
Feature activation+0.000
ile
Tokenile
Feature activation+0.000
(
Token (
Feature activation+0.000
Michael
TokenMichael
Feature activation+0.000
C
Token C
Feature activation+0.000
aine
Tokenaine
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
ile
Tokenile
Feature activation+0.000
(
Token (
Feature activation+0.000
Michael
TokenMichael
Feature activation+0.000
C
Token C
Feature activation+0.000
aine
Tokenaine
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
(
Token (
Feature activation+0.000
Michael
TokenMichael
Feature activation+0.000
C
Token C
Feature activation+0.000
aine
Tokenaine
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Michael
TokenMichael
Feature activation+0.000
C
Token C
Feature activation+0.000
aine
Tokenaine
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
anking
Tokenanking
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
anking
Tokenanking
Feature activation+0.000
the
Token the
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
C
Token C
Feature activation+0.000
aine
Tokenaine
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
aine
Tokenaine
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000

Top DFA by src position
MAX = 0.188

regulated
Token regulated
Feature activation-0.000
Top resid features:
by
Token by
Feature activation+0.023
Top resid features:
the
Token the
Feature activation+0.024
Top resid features:
residues
Token residues
Feature activation+0.003
Top resid features:
fl
Token fl
Feature activation+0.047
Top resid features:
anking
Tokenanking
Feature activation+0.105
Top resid features:
the
Token the
Feature activation+0.000
Top resid features:
sc
Token sc
Feature activation+0.000
Top resid features:
iss
Tokeniss
Feature activation+0.000
Top resid features:
ile
Tokenile
Feature activation+0.000
Top resid features:
bonds
Token bonds
Feature activation+0.000
Top resid features:
G
Token G
Feature activation+0.016
Top resid features:
ag
Tokenag
Feature activation-0.011
Top resid features:
is
Token is
Feature activation+0.018
Top resid features:
regulated
Token regulated
Feature activation+0.000
Top resid features:
by
Token by
Feature activation+0.052
Top resid features:
the
Token the
Feature activation+0.057
Top resid features:
residues
Token residues
Feature activation-0.025
Top resid features:
fl
Token fl
Feature activation+0.055
Top resid features:
anking
Tokenanking
Feature activation+0.000
Top resid features:
the
Token the
Feature activation+0.000
Top resid features:
sc
Token sc
Feature activation+0.000
Top resid features:
G
Token G
Feature activation+0.015
Top resid features:
ag
Tokenag
Feature activation-0.008
Top resid features:
is
Token is
Feature activation+0.073
Top resid features:
regulated
Token regulated
Feature activation-0.020
Top resid features:
by
Token by
Feature activation+0.082
Top resid features:
the
Token the
Feature activation+0.129
Top resid features:
residues
Token residues
Feature activation+0.000
Top resid features:
fl
Token fl
Feature activation+0.000
Top resid features:
anking
Tokenanking
Feature activation+0.000
Top resid features:
the
Token the
Feature activation+0.000
Top resid features:
sc
Token sc
Feature activation+0.000
Top resid features:
of
Token of
Feature activation+0.020
Top resid features:
G
Token G
Feature activation+0.018
Top resid features:
ag
Tokenag
Feature activation-0.007
Top resid features:
is
Token is
Feature activation+0.025
Top resid features:
regulated
Token regulated
Feature activation-0.014
Top resid features:
by
Token by
Feature activation+0.055
Top resid features:
the
Token the
Feature activation+0.053
Top resid features:
residues
Token residues
Feature activation-0.061
Top resid features:
fl
Token fl
Feature activation+0.000
Top resid features:
anking
Tokenanking
Feature activation+0.000
Top resid features:
the
Token the
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.019
Top resid features:
ile
Tokenile
Feature activation-0.029
Top resid features:
(
Token (
Feature activation+0.115
Top resid features:
Michael
TokenMichael
Feature activation-0.020
Top resid features:
C
Token C
Feature activation+0.035
Top resid features:
aine
Tokenaine
Feature activation-0.041
Top resid features:
)
Token)
Feature activation-0.062
Top resid features:
in
Token in
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.041
Top resid features:
ile
Tokenile
Feature activation-0.048
Top resid features:
(
Token (
Feature activation+0.079
Top resid features:
Michael
TokenMichael
Feature activation+0.002
Top resid features:
C
Token C
Feature activation+0.028
Top resid features:
aine
Tokenaine
Feature activation+0.097
Top resid features:
)
Token)
Feature activation+0.000
Top resid features:
in
Token in
Feature activation+0.000
Top resid features:
the
Token the
Feature activation+0.000
Top resid features:
process
Token process
Feature activation+0.000
Top resid features:
.
Token.
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.024
Top resid features:
ile
Tokenile
Feature activation-0.018
Top resid features:
(
Token (
Feature activation+0.089
Top resid features:
Michael
TokenMichael
Feature activation-0.036
Top resid features:
C
Token C
Feature activation+0.021
Top resid features:
aine
Tokenaine
Feature activation-0.035
Top resid features:
)
Token)
Feature activation-0.031
Top resid features:
in
Token in
Feature activation+0.036
Top resid features:
Michael
TokenMichael
Feature activation-0.053
Top resid features:
C
Token C
Feature activation+0.009
Top resid features:
aine
Tokenaine
Feature activation-0.040
Top resid features:
)
Token)
Feature activation-0.034
Top resid features:
in
Token in
Feature activation+0.064
Top resid features:
the
Token the
Feature activation+0.149
Top resid features:
process
Token process
Feature activation+0.000
Top resid features:
.
Token.
Feature activation+0.000
Top resid features:
Ċ
TokenĊ
Feature activation+0.000
Top resid features:
Ċ
TokenĊ
Feature activation+0.000
Top resid features:
The
TokenThe
Feature activation+0.000
Top resid features:
the
Token the
Feature activation+0.019
Top resid features:
process
Token process
Feature activation+0.011
Top resid features:
.
Token.
Feature activation+0.019
Top resid features:
Ċ
TokenĊ
Feature activation-0.011
Top resid features:
Ċ
TokenĊ
Feature activation-0.009
Top resid features:
The
TokenThe
Feature activation+0.059
Top resid features:
processing
Token processing
Feature activation-0.007
Top resid features:
of
Token of
Feature activation+0.045
Top resid features:
G
Token G
Feature activation+0.049
Top resid features:
ag
Tokenag
Feature activation+0.000
Top resid features:
is
Token is
Feature activation+0.000
Top resid features:
the
Token the
Feature activation+0.022
Top resid features:
process
Token process
Feature activation+0.017
Top resid features:
.
Token.
Feature activation+0.046
Top resid features:
Ċ
TokenĊ
Feature activation+0.006
Top resid features:
Ċ
TokenĊ
Feature activation+0.010
Top resid features:
The
TokenThe
Feature activation+0.078
Top resid features:
processing
Token processing
Feature activation-0.060
Top resid features:
of
Token of
Feature activation+0.048
Top resid features:
G
Token G
Feature activation+0.000
Top resid features:
ag
Tokenag
Feature activation+0.000
Top resid features:
is
Token is
Feature activation+0.000
Top resid features:
aine
Tokenaine
Feature activation-0.028
Top resid features:
)
Token)
Feature activation-0.028
Top resid features:
in
Token in
Feature activation+0.016
Top resid features:
the
Token the
Feature activation+0.035
Top resid features:
process
Token process
Feature activation+0.009
Top resid features:
.
Token.
Feature activation+0.078
Top resid features:
Ċ
TokenĊ
Feature activation+0.010
Top resid features:
Ċ
TokenĊ
Feature activation+0.015
Top resid features:
The
TokenThe
Feature activation+0.059
Top resid features:
processing
Token processing
Feature activation+0.000
Top resid features:
of
Token of
Feature activation+0.000
Top resid features:
the
Token the
Feature activation+0.014
Top resid features:
process
Token process
Feature activation+0.010
Top resid features:
.
Token.
Feature activation+0.015
Top resid features:
Ċ
TokenĊ
Feature activation-0.004
Top resid features:
Ċ
TokenĊ
Feature activation-0.001
Top resid features:
The
TokenThe
Feature activation+0.087
Top resid features:
processing
Token processing
Feature activation-0.129
Top resid features:
of
Token of
Feature activation+0.000
Top resid features:
G
Token G
Feature activation+0.000
Top resid features:
ag
Tokenag
Feature activation+0.000
Top resid features:
is
Token is
Feature activation+0.000
Top resid features:
The
TokenThe
Feature activation+0.053
Top resid features:
processing
Token processing
Feature activation-0.017
Top resid features:
of
Token of
Feature activation+0.030
Top resid features:
G
Token G
Feature activation+0.012
Top resid features:
ag
Tokenag
Feature activation-0.011
Top resid features:
is
Token is
Feature activation+0.188
Top resid features:
regulated
Token regulated
Feature activation+0.000
Top resid features:
by
Token by
Feature activation+0.000
Top resid features:
the
Token the
Feature activation+0.000
Top resid features:
residues
Token residues
Feature activation+0.000
Top resid features:
fl
Token fl
Feature activation+0.000
Top resid features:
.
Token.
Feature activation+0.033
Top resid features:
Ċ
TokenĊ
Feature activation-0.009
Top resid features:
Ċ
TokenĊ
Feature activation-0.007
Top resid features:
The
TokenThe
Feature activation+0.037
Top resid features:
processing
Token processing
Feature activation+0.000
Top resid features:
of
Token of
Feature activation+0.043
Top resid features:
G
Token G
Feature activation+0.034
Top resid features:
ag
Tokenag
Feature activation-0.098
Top resid features:
is
Token is
Feature activation+0.000
Top resid features:
regulated
Token regulated
Feature activation+0.000
Top resid features:
by
Token by
Feature activation+0.000
Top resid features:
The
TokenThe
Feature activation+0.036
Top resid features:
processing
Token processing
Feature activation+0.004
Top resid features:
of
Token of
Feature activation+0.040
Top resid features:
G
Token G
Feature activation+0.016
Top resid features:
ag
Tokenag
Feature activation+0.000
Top resid features:
is
Token is
Feature activation+0.079
Top resid features:
regulated
Token regulated
Feature activation-0.001
Top resid features:
by
Token by
Feature activation+0.000
Top resid features:
the
Token the
Feature activation+0.000
Top resid features:
residues
Token residues
Feature activation+0.000
Top resid features:
fl
Token fl
Feature activation+0.000
Top resid features:
The
TokenThe
Feature activation+0.040
Top resid features:
processing
Token processing
Feature activation-0.005
Top resid features:
of
Token of
Feature activation+0.030
Top resid features:
G
Token G
Feature activation+0.017
Top resid features:
ag
Tokenag
Feature activation-0.013
Top resid features:
is
Token is
Feature activation+0.068
Top resid features:
regulated
Token regulated
Feature activation+0.012
Top resid features:
by
Token by
Feature activation-0.001
Top resid features:
the
Token the
Feature activation+0.000
Top resid features:
residues
Token residues
Feature activation+0.000
Top resid features:
fl
Token fl
Feature activation+0.000
Top resid features:
aine
Tokenaine
Feature activation-0.041
Top resid features:
)
Token)
Feature activation-0.036
Top resid features:
in
Token in
Feature activation+0.015
Top resid features:
the
Token the
Feature activation+0.036
Top resid features:
process
Token process
Feature activation+0.012
Top resid features:
.
Token.
Feature activation+0.112
Top resid features:
Ċ
TokenĊ
Feature activation-0.035
Top resid features:
Ċ
TokenĊ
Feature activation-0.022
Top resid features:
The
TokenThe
Feature activation+0.000
Top resid features:
processing
Token processing
Feature activation+0.000
Top resid features:
of
Token of
Feature activation+0.000
Top resid features:
aine
Tokenaine
Feature activation-0.041
Top resid features:
)
Token)
Feature activation-0.039
Top resid features:
in
Token in
Feature activation+0.019
Top resid features:
the
Token the
Feature activation+0.048
Top resid features:
process
Token process
Feature activation+0.016
Top resid features:
.
Token.
Feature activation+0.152
Top resid features:
Ċ
TokenĊ
Feature activation-0.057
Top resid features:
Ċ
TokenĊ
Feature activation+0.000
Top resid features:
The
TokenThe
Feature activation+0.000
Top resid features:
processing
Token processing
Feature activation+0.000
Top resid features:
of
Token of
Feature activation+0.000
Top resid features:
Michael
TokenMichael
Feature activation-0.018
Top resid features:
C
Token C
Feature activation+0.024
Top resid features:
aine
Tokenaine
Feature activation-0.025
Top resid features:
)
Token)
Feature activation-0.026
Top resid features:
in
Token in
Feature activation+0.046
Top resid features:
the
Token the
Feature activation+0.095
Top resid features:
process
Token process
Feature activation+0.078
Top resid features:
.
Token.
Feature activation+0.000
Top resid features:
Ċ
TokenĊ
Feature activation+0.000
Top resid features:
Ċ
TokenĊ
Feature activation+0.000
Top resid features:
The
TokenThe
Feature activation+0.000
Top resid features:
aine
Tokenaine
Feature activation-0.033
Top resid features:
)
Token)
Feature activation-0.045
Top resid features:
in
Token in
Feature activation+0.016
Top resid features:
the
Token the
Feature activation+0.056
Top resid features:
process
Token process
Feature activation+0.035
Top resid features:
.
Token.
Feature activation+0.102
Top resid features:
Ċ
TokenĊ
Feature activation+0.000
Top resid features:
Ċ
TokenĊ
Feature activation+0.000
Top resid features:
The
TokenThe
Feature activation+0.000
Top resid features:
processing
Token processing
Feature activation+0.000
Top resid features:
of
Token of
Feature activation+0.000
Top resid features:

Decoder Weights Distribution

Head 0: 0.07

Head 1: 0.09

Head 2: 0.07

Head 3: 0.07

Head 4: 0.08

Head 5: 0.07

Head 6: 0.07

Head 7: 0.09

Head 8: 0.07

Head 9: 0.10

Head 10: 0.08

Head 11: 0.14

Positive logits

olulu4.50

��4.20

undai3.89

ensing3.52

psychiat3.27

pregn3.13

unden3.11

ividual3.08

ribune3.08

uckland3.05

suspic3.00

judicial2.99

��2.95

anmar2.95

unal2.94

uncture2.94

xual2.92

iminary2.91

terday2.89

��2.88

Negative logits

estern-3.12

-2.89

Weasley-2.86

dor-2.62

bars-2.61

Cod-2.61

haus-2.58

ergic-2.56

ーク-2.51

DOS-2.50

minster-2.50

ーテ-2.49

sels-2.47

qua-2.46

boats-2.45

mania-2.45

loe-2.43

cough-2.43

Voyager-2.43

ford-2.40

INTERVAL 0.000 - 0.000
CONTAINS 100.000%

und
Tokenund
Feature activation+0.000
rums
Tokenrums
Feature activation+0.000
on
Token on
Feature activation+0.000
Facebook
Token Facebook
Feature activation+0.000
here
Token here
Feature activation+0.000
.
Token.
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
1
Token1
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
right
Token right
Feature activation+0.000
to
Token to
Feature activation+0.000
left
Token left
Feature activation+0.000
.
Token.
Feature activation+0.000
Later
Token Later
Feature activation+0.000
,
Token,
Feature activation+0.000
this
Token this
Feature activation+0.000
evolved
Token evolved
Feature activation+0.000
to
Token to
Feature activation+0.000
âĢ
Token âĢ
Feature activation+0.000
ľ
Tokenľ
Feature activation+0.000
-
Token-
Feature activation+0.000
time
Tokentime
Feature activation+0.000
said
Token said
Feature activation+0.000
they
Token they
Feature activation+0.000
opted
Token opted
Feature activation+0.000
for
Token for
Feature activation+0.000
part
Token part
Feature activation+0.000
-
Token-
Feature activation+0.000
time
Tokentime
Feature activation+0.000
work
Token work
Feature activation+0.000
because
Token because
Feature activation+0.000
earth
Token earth
Feature activation+0.000
has
Token has
Feature activation+0.000
been
Token been
Feature activation+0.000
visited
Token visited
Feature activation+0.000
is
Token is
Feature activation+0.000
found
Token found
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
For
TokenFor
Feature activation+0.000
more
Token more
Feature activation+0.000
than
Token than
Feature activation+0.000
50
Token 50
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
He
TokenHe
Feature activation+0.000
stood
Token stood
Feature activation+0.000
sto
Token sto
Feature activation+0.000
ically
Tokenically
Feature activation+0.000
as
Token as
Feature activation+0.000
the
Token the
Feature activation+0.000
verdict
Token verdict
Feature activation+0.000
was
Token was
Feature activation+0.000
read
Token read
Feature activation+0.000
,
Token,
Feature activation+0.000

BOTTOM ACTIVATIONS
MIN = 0.000

regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
anking
Tokenanking
Feature activation+0.000
the
Token the
Feature activation+0.000
sc
Token sc
Feature activation+0.000
iss
Tokeniss
Feature activation+0.000
ile
Tokenile
Feature activation+0.000
bonds
Token bonds
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
anking
Tokenanking
Feature activation+0.000
the
Token the
Feature activation+0.000
sc
Token sc
Feature activation+0.000
iss
Tokeniss
Feature activation+0.000
ile
Tokenile
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
anking
Tokenanking
Feature activation+0.000
the
Token the
Feature activation+0.000
sc
Token sc
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
anking
Tokenanking
Feature activation+0.000
the
Token the
Feature activation+0.000
sc
Token sc
Feature activation+0.000
iss
Tokeniss
Feature activation+0.000
ile
Tokenile
Feature activation+0.000
(
Token (
Feature activation+0.000
Michael
TokenMichael
Feature activation+0.000
C
Token C
Feature activation+0.000
aine
Tokenaine
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
ile
Tokenile
Feature activation+0.000
(
Token (
Feature activation+0.000
Michael
TokenMichael
Feature activation+0.000
C
Token C
Feature activation+0.000
aine
Tokenaine
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
(
Token (
Feature activation+0.000
Michael
TokenMichael
Feature activation+0.000
C
Token C
Feature activation+0.000
aine
Tokenaine
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Michael
TokenMichael
Feature activation+0.000
C
Token C
Feature activation+0.000
aine
Tokenaine
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
anking
Tokenanking
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
anking
Tokenanking
Feature activation+0.000
the
Token the
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
C
Token C
Feature activation+0.000
aine
Tokenaine
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
aine
Tokenaine
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000

Top feature 5 in H0.11: (feature 5334

TOP ACTIVATIONS
MAX = 4.034

<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
bron
Token bron
Feature activation+0.000
co
Tokenco
Feature activation+0.000
requiring
Token requiring
Feature activation+0.000
strength
Token strength
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+1.969
Black
TokenBlack
Feature activation+0.000
Lives
Token Lives
Feature activation+0.000
Matter
Token Matter
Feature activation+0.000
.
Token.
Feature activation+0.000
All
Token All
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
W
TokenW
Feature activation+0.000
is
Tokenis
Feature activation+0.000
eman
Tokeneman
Feature activation+0.000
AP
TokenAP
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+1.905
Air
TokenAir
Feature activation+0.000
guns
Token guns
Feature activation+0.000
used
Token used
Feature activation+0.000
for
Token for
Feature activation+0.000
marine
Token marine
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
m
Tokenm
Feature activation+0.000
gonna
Token gonna
Feature activation+0.000
call
Token call
Feature activation+0.000
Ted
Token Ted
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+1.691
Posted
TokenPosted
Feature activation+0.000
in
Token in
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.055
Ċ
TokenĊ
Feature activation+0.000
Dear
TokenDear
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
izers
Tokenizers
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ŀ
TokenĿ
Feature activation+0.000
âĢĵ
Token âĢĵ
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+1.404
Due
TokenDue
Feature activation+0.000
to
Token to
Feature activation+0.000
boring
Token boring
Feature activation+0.000
circumstances
Token circumstances
Feature activation+0.000
beyond
Token beyond
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
Like
Token Like
Feature activation+0.000
Loading
Token Loading
Feature activation+0.000
...
Token...
Feature activation+0.000
Related
Token Related
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+1.291
5
Token5
Feature activation+0.000
.
Token.
Feature activation+0.000
0
Token0
Feature activation+0.000
âĺħ
Token âĺħ
Feature activation+0.000
âĺħâĺħ
Tokenâĺħâĺħ
Feature activation+0.000
te
Tokente
Feature activation+0.000
al
Tokenal
Feature activation+0.000
ane
Tokenane
Feature activation+0.000
z
Tokenz
Feature activation+0.000
@
Token@
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+1.246
H
TokenH
Feature activation+0.000
ollywood
Tokenollywood
Feature activation+0.000
's
Token's
Feature activation+0.000
highest
Token highest
Feature activation+0.000
profile
Token profile
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
ofi
Tokenofi
Feature activation+0.000
was
Token was
Feature activation+0.000
spectacular
Token spectacular
Feature activation+0.000
and
Token and
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+1.188
Black
TokenBlack
Feature activation+0.000
Girls
Token Girls
Feature activation+0.000
Rock
Token Rock
Feature activation+0.000
!
Token!
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
angular
Token angular
Feature activation+0.000
js
Tokenjs
Feature activation+0.000
etc
Token etc
Feature activation+0.000
and
Token and
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+1.168
Q
TokenQ
Feature activation+0.000
atar
Tokenatar
Feature activation+0.000
are
Token are
Feature activation+0.000
set
Token set
Feature activation+0.000
to
Token to
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
of
Token of
Feature activation+0.000
religious
Token religious
Feature activation+0.000
groups
Token groups
Feature activation+0.000
and
Token and
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+1.132
SH
TokenSH
Feature activation+0.000
OW
TokenOW
Feature activation+0.000
US
Token US
Feature activation+0.000
DET
Token DET
Feature activation+0.000
RO
TokenRO
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
sc
Tokensc
Feature activation+0.000
ast
Tokenast
Feature activation+0.000
unit
Token unit
Feature activation+0.000
:
Token:
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+1.066
Wars
TokenWars
Feature activation+0.000
were
Token were
Feature activation+0.000
fought
Token fought
Feature activation+0.000
to
Token to
Feature activation+0.000
impose
Token impose
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
the
Token the
Feature activation+0.000
powerful
Token powerful
Feature activation+0.000
incentive
Token incentive
Feature activation+0.000
of
Token of
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+1.063
Police
TokenPolice
Feature activation+0.000
snap
Token snap
Feature activation+0.000
up
Token up
Feature activation+0.000
mud
Token mud
Feature activation+0.000
crab
Token crab
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
band
Token band
Feature activation+0.000
ing
Tokening
Feature activation+0.000
together
Token together
Feature activation+0.000
,
Token,
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+1.012
After
TokenAfter
Feature activation+0.000
talking
Token talking
Feature activation+0.000
about
Token about
Feature activation+0.000
how
Token how
Feature activation+0.000
unlikely
Token unlikely
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
:
Token:
Feature activation+0.000
1993
Token 1993
Feature activation+0.000
âĢĵ
TokenâĢĵ
Feature activation+0.000
1995
Token1995
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.952
Rel
TokenRel
Feature activation+0.000
igious
Tokenigious
Feature activation+0.000
leaders
Token leaders
Feature activation+0.000
,
Token,
Feature activation+0.000
including
Token including
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
the
Token the
Feature activation+0.000
number
Token number
Feature activation+0.000
of
Token of
Feature activation+0.000
highly
Token highly
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.841
MS
TokenMS
Feature activation+0.000
NBC
TokenNBC
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
s
Tokens
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
for
Token for
Feature activation+0.000
the
Token the
Feature activation+0.000
future
Token future
Feature activation+0.000
."
Token."
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.788
DON
TokenDON
Feature activation+0.000
ALD
TokenALD
Feature activation+0.000
Trump
Token Trump
Feature activation+0.000
wasn
Token wasn
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
initially
Token initially
Feature activation+0.000
returned
Token returned
Feature activation+0.000
Wednesday
Token Wednesday
Feature activation+0.000
.
Token.
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.695
A
TokenA
Feature activation+0.000
team
Token team
Feature activation+0.000
led
Token led
Feature activation+0.000
by
Token by
Feature activation+0.000
post
Token post
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
to
Token to
Feature activation+0.000
this
Token this
Feature activation+0.000
report
Token report
Feature activation+0.000
.
Token.
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.671
No
TokenNo
Feature activation+0.000
self
Token self
Feature activation+0.000
-
Token-
Feature activation+0.000
respect
Tokenrespect
Feature activation+0.000
ing
Tokening
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
@
Token@
Feature activation+0.000
gmail
Tokengmail
Feature activation+0.000
.
Token.
Feature activation+0.000
com
Tokencom
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.671
Keith
TokenKeith
Feature activation+0.000
Law
Token Law
Feature activation+0.000
gets
Token gets
Feature activation+0.000
Twitter
Token Twitter
Feature activation+0.000
suspension
Token suspension
Feature activation+0.000
shift
Token shift
Feature activation+0.000
in
Token in
Feature activation+0.000
consciousness
Token consciousness
Feature activation+0.000
on
Token on
Feature activation+0.000
this
Token this
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.665
In
TokenIn
Feature activation+0.000
keeping
Token keeping
Feature activation+0.000
with
Token with
Feature activation+0.000
holiday
Token holiday
Feature activation+0.000
tradition
Token tradition
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
Senator
Token Senator
Feature activation+0.000
Ted
Token Ted
Feature activation+0.000
Cruz
Token Cruz
Feature activation+0.000
.
Token.
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.575
Anthony
TokenAnthony
Feature activation+0.000
Log
Token Log
Feature activation+0.000
istics
Tokenistics
Feature activation+0.000
for
Token for
Feature activation+0.000
Men
Token Men
Feature activation+0.000

Top DFA by src position
MAX = 8.257

<|endoftext|>
Token<|endoftext|>
Feature activation+2.482
Top resid features:
bron
Token bron
Feature activation+0.128
Top resid features:
co
Tokenco
Feature activation+0.097
Top resid features:
requiring
Token requiring
Feature activation+0.164
Top resid features:
strength
Token strength
Feature activation+0.139
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+8.132
Top resid features:
Black
TokenBlack
Feature activation+0.000
Top resid features:
Lives
Token Lives
Feature activation+0.000
Top resid features:
Matter
Token Matter
Feature activation+0.000
Top resid features:
.
Token.
Feature activation+0.000
Top resid features:
All
Token All
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+2.495
Top resid features:
W
TokenW
Feature activation+0.125
Top resid features:
is
Tokenis
Feature activation+0.127
Top resid features:
eman
Tokeneman
Feature activation+0.075
Top resid features:
AP
TokenAP
Feature activation-0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+8.257
Top resid features:
Air
TokenAir
Feature activation+0.000
Top resid features:
guns
Token guns
Feature activation+0.000
Top resid features:
used
Token used
Feature activation+0.000
Top resid features:
for
Token for
Feature activation+0.000
Top resid features:
marine
Token marine
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+2.409
Top resid features:
m
Tokenm
Feature activation+0.076
Top resid features:
gonna
Token gonna
Feature activation+0.178
Top resid features:
call
Token call
Feature activation+0.019
Top resid features:
Ted
Token Ted
Feature activation+0.014
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+8.167
Top resid features:
Posted
TokenPosted
Feature activation+0.000
Top resid features:
in
Token in
Feature activation+0.000
Top resid features:
Ċ
TokenĊ
Feature activation+0.000
Top resid features:
Ċ
TokenĊ
Feature activation+0.000
Top resid features:
Dear
TokenDear
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+2.358
Top resid features:
izers
Tokenizers
Feature activation-0.089
Top resid features:
âĢ
TokenâĢ
Feature activation+0.091
Top resid features:
Ŀ
TokenĿ
Feature activation+0.039
Top resid features:
âĢĵ
Token âĢĵ
Feature activation+0.178
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+7.999
Top resid features:
Due
TokenDue
Feature activation+0.000
Top resid features:
to
Token to
Feature activation+0.000
Top resid features:
boring
Token boring
Feature activation+0.000
Top resid features:
circumstances
Token circumstances
Feature activation+0.000
Top resid features:
beyond
Token beyond
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+2.151
Top resid features:
Like
Token Like
Feature activation+0.019
Top resid features:
Loading
Token Loading
Feature activation+0.280
Top resid features:
...
Token...
Feature activation+0.122
Top resid features:
Related
Token Related
Feature activation+0.184
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+7.707
Top resid features:
5
Token5
Feature activation+0.000
Top resid features:
.
Token.
Feature activation+0.000
Top resid features:
0
Token0
Feature activation+0.000
Top resid features:
âĺħ
Token âĺħ
Feature activation+0.000
Top resid features:
âĺħâĺħ
Tokenâĺħâĺħ
Feature activation+0.000
Top resid features:
te
Tokente
Feature activation+0.038
Top resid features:
al
Tokenal
Feature activation-0.085
Top resid features:
ane
Tokenane
Feature activation+0.025
Top resid features:
z
Tokenz
Feature activation+0.009
Top resid features:
@
Token@
Feature activation+0.073
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+7.979
Top resid features:
H
TokenH
Feature activation+0.000
Top resid features:
ollywood
Tokenollywood
Feature activation+0.000
Top resid features:
's
Token's
Feature activation+0.000
Top resid features:
highest
Token highest
Feature activation+0.000
Top resid features:
profile
Token profile
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+2.052
Top resid features:
ofi
Tokenofi
Feature activation+0.083
Top resid features:
was
Token was
Feature activation+0.075
Top resid features:
spectacular
Token spectacular
Feature activation+0.275
Top resid features:
and
Token and
Feature activation+0.118
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+7.757
Top resid features:
Black
TokenBlack
Feature activation+0.000
Top resid features:
Girls
Token Girls
Feature activation+0.000
Top resid features:
Rock
Token Rock
Feature activation+0.000
Top resid features:
!
Token!
Feature activation+0.000
Top resid features:
Ċ
TokenĊ
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+2.149
Top resid features:
angular
Token angular
Feature activation+0.098
Top resid features:
js
Tokenjs
Feature activation+0.069
Top resid features:
etc
Token etc
Feature activation+0.143
Top resid features:
and
Token and
Feature activation+0.082
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+7.799
Top resid features:
Q
TokenQ
Feature activation+0.000
Top resid features:
atar
Tokenatar
Feature activation+0.000
Top resid features:
are
Token are
Feature activation+0.000
Top resid features:
set
Token set
Feature activation+0.000
Top resid features:
to
Token to
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+2.038
Top resid features:
of
Token of
Feature activation-0.002
Top resid features:
religious
Token religious
Feature activation+0.236
Top resid features:
groups
Token groups
Feature activation+0.137
Top resid features:
and
Token and
Feature activation+0.146
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+7.749
Top resid features:
SH
TokenSH
Feature activation+0.000
Top resid features:
OW
TokenOW
Feature activation+0.000
Top resid features:
US
Token US
Feature activation+0.000
Top resid features:
DET
Token DET
Feature activation+0.000
Top resid features:
RO
TokenRO
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+2.230
Top resid features:
sc
Tokensc
Feature activation+0.001
Top resid features:
ast
Tokenast
Feature activation-0.052
Top resid features:
unit
Token unit
Feature activation+0.154
Top resid features:
:
Token:
Feature activation-0.105
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+8.010
Top resid features:
Wars
TokenWars
Feature activation+0.000
Top resid features:
were
Token were
Feature activation+0.000
Top resid features:
fought
Token fought
Feature activation+0.000
Top resid features:
to
Token to
Feature activation+0.000
Top resid features:
impose
Token impose
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+1.935
Top resid features:
the
Token the
Feature activation+0.069
Top resid features:
powerful
Token powerful
Feature activation+0.177
Top resid features:
incentive
Token incentive
Feature activation+0.181
Top resid features:
of
Token of
Feature activation+0.089
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+7.784
Top resid features:
Police
TokenPolice
Feature activation+0.000
Top resid features:
snap
Token snap
Feature activation+0.000
Top resid features:
up
Token up
Feature activation+0.000
Top resid features:
mud
Token mud
Feature activation+0.000
Top resid features:
crab
Token crab
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+2.138
Top resid features:
band
Token band
Feature activation+0.071
Top resid features:
ing
Tokening
Feature activation+0.030
Top resid features:
together
Token together
Feature activation+0.067
Top resid features:
,
Token,
Feature activation-0.038
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+7.916
Top resid features:
After
TokenAfter
Feature activation+0.000
Top resid features:
talking
Token talking
Feature activation+0.000
Top resid features:
about
Token about
Feature activation+0.000
Top resid features:
how
Token how
Feature activation+0.000
Top resid features:
unlikely
Token unlikely
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+2.080
Top resid features:
:
Token:
Feature activation+0.003
Top resid features:
1993
Token 1993
Feature activation+0.110
Top resid features:
âĢĵ
TokenâĢĵ
Feature activation+0.129
Top resid features:
1995
Token1995
Feature activation+0.047
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+7.755
Top resid features:
Rel
TokenRel
Feature activation+0.000
Top resid features:
igious
Tokenigious
Feature activation+0.000
Top resid features:
leaders
Token leaders
Feature activation+0.000
Top resid features:
,
Token,
Feature activation+0.000
Top resid features:
including
Token including
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+1.965
Top resid features:
the
Token the
Feature activation+0.067
Top resid features:
number
Token number
Feature activation+0.120
Top resid features:
of
Token of
Feature activation+0.058
Top resid features:
highly
Token highly
Feature activation+0.055
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+7.749
Top resid features:
MS
TokenMS
Feature activation+0.000
Top resid features:
NBC
TokenNBC
Feature activation+0.000
Top resid features:
âĢ
TokenâĢ
Feature activation+0.000
Top resid features:
Ļ
TokenĻ
Feature activation+0.000
Top resid features:
s
Tokens
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+1.974
Top resid features:
for
Token for
Feature activation+0.134
Top resid features:
the
Token the
Feature activation-0.043
Top resid features:
future
Token future
Feature activation+0.116
Top resid features:
."
Token."
Feature activation+0.124
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+7.657
Top resid features:
DON
TokenDON
Feature activation+0.000
Top resid features:
ALD
TokenALD
Feature activation+0.000
Top resid features:
Trump
Token Trump
Feature activation+0.000
Top resid features:
wasn
Token wasn
Feature activation+0.000
Top resid features:
âĢ
TokenâĢ
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+2.014
Top resid features:
initially
Token initially
Feature activation+0.116
Top resid features:
returned
Token returned
Feature activation+0.048
Top resid features:
Wednesday
Token Wednesday
Feature activation+0.106
Top resid features:
.
Token.
Feature activation-0.206
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+7.788
Top resid features:
A
TokenA
Feature activation+0.000
Top resid features:
team
Token team
Feature activation+0.000
Top resid features:
led
Token led
Feature activation+0.000
Top resid features:
by
Token by
Feature activation+0.000
Top resid features:
post
Token post
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+1.933
Top resid features:
to
Token to
Feature activation+0.128
Top resid features:
this
Token this
Feature activation+0.112
Top resid features:
report
Token report
Feature activation+0.110
Top resid features:
.
Token.
Feature activation-0.054
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+7.615
Top resid features:
No
TokenNo
Feature activation+0.000
Top resid features:
self
Token self
Feature activation+0.000
Top resid features:
-
Token-
Feature activation+0.000
Top resid features:
respect
Tokenrespect
Feature activation+0.000
Top resid features:
ing
Tokening
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+2.038
Top resid features:
@
Token@
Feature activation+0.077
Top resid features:
gmail
Tokengmail
Feature activation+0.121
Top resid features:
.
Token.
Feature activation-0.209
Top resid features:
com
Tokencom
Feature activation+0.113
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+7.704
Top resid features:
Keith
TokenKeith
Feature activation+0.000
Top resid features:
Law
Token Law
Feature activation+0.000
Top resid features:
gets
Token gets
Feature activation+0.000
Top resid features:
Twitter
Token Twitter
Feature activation+0.000
Top resid features:
suspension
Token suspension
Feature activation+0.000
Top resid features:
shift
Token shift
Feature activation+0.026
Top resid features:
in
Token in
Feature activation+0.127
Top resid features:
consciousness
Token consciousness
Feature activation+0.082
Top resid features:
on
Token on
Feature activation+0.147
Top resid features:
this
Token this
Feature activation+0.105
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+7.420
Top resid features:
In
TokenIn
Feature activation+0.000
Top resid features:
keeping
Token keeping
Feature activation+0.000
Top resid features:
with
Token with
Feature activation+0.000
Top resid features:
holiday
Token holiday
Feature activation+0.000
Top resid features:
tradition
Token tradition
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+2.014
Top resid features:
Senator
Token Senator
Feature activation-0.060
Top resid features:
Ted
Token Ted
Feature activation+0.094
Top resid features:
Cruz
Token Cruz
Feature activation+0.104
Top resid features:
.
Token.
Feature activation-0.209
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+7.805
Top resid features:
Anthony
TokenAnthony
Feature activation+0.000
Top resid features:
Log
Token Log
Feature activation+0.000
Top resid features:
istics
Tokenistics
Feature activation+0.000
Top resid features:
for
Token for
Feature activation+0.000
Top resid features:
Men
Token Men
Feature activation+0.000
Top resid features:

Decoder Weights Distribution

Head 0: 0.06

Head 1: 0.09

Head 2: 0.06

Head 3: 0.08

Head 4: 0.07

Head 5: 0.08

Head 6: 0.10

Head 7: 0.07

Head 8: 0.05

Head 9: 0.14

Head 10: 0.07

Head 11: 0.14

Positive logits

▬▬3.33

ONSORED2.81

NESS2.71

Products2.71

pmwiki2.71

Interested2.69

LOS2.69

Caption2.65

VIDEOS2.54

advertisement2.53

WASHINGTON2.51

NetMessage2.50

FIELD2.45

CLOSE2.45

wcsstore2.44

Synopsis2.42

Surviv2.41

UPDATE2.40

Temperature2.36

Product2.35

Negative logits

deported-2.68

��-2.61

tradem-2.20

boarded-2.19

adra-2.15

migration-2.13

refugees-2.10

illions-2.08

sailed-2.03

asel-2.03

allied-2.02

pheus-2.01

councill-2.01

thous-2.00

hunted-1.99

surrendered-1.99

uese-1.97

thora-1.94

afterwards-1.93

afterward-1.93

INTERVAL 3.631 - 4.034
CONTAINS 0.001%

INTERVAL 3.227 - 3.631
CONTAINS 0.001%

INTERVAL 2.824 - 3.227
CONTAINS 0.001%

INTERVAL 2.421 - 2.824
CONTAINS 0.000%

INTERVAL 2.017 - 2.421
CONTAINS 0.001%

INTERVAL 1.614 - 2.017
CONTAINS 0.001%

<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
W
TokenW
Feature activation+0.000
is
Tokenis
Feature activation+0.000
eman
Tokeneman
Feature activation+0.000
AP
TokenAP
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+1.905
Air
TokenAir
Feature activation+0.000
guns
Token guns
Feature activation+0.000
used
Token used
Feature activation+0.000
for
Token for
Feature activation+0.000
marine
Token marine
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
bron
Token bron
Feature activation+0.000
co
Tokenco
Feature activation+0.000
requiring
Token requiring
Feature activation+0.000
strength
Token strength
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+1.969
Black
TokenBlack
Feature activation+0.000
Lives
Token Lives
Feature activation+0.000
Matter
Token Matter
Feature activation+0.000
.
Token.
Feature activation+0.000
All
Token All
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
m
Tokenm
Feature activation+0.000
gonna
Token gonna
Feature activation+0.000
call
Token call
Feature activation+0.000
Ted
Token Ted
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+1.691
Posted
TokenPosted
Feature activation+0.000
in
Token in
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.055
Ċ
TokenĊ
Feature activation+0.000
Dear
TokenDear
Feature activation+0.000

INTERVAL 1.210 - 1.614
CONTAINS 0.001%

te
Tokente
Feature activation+0.000
al
Tokenal
Feature activation+0.000
ane
Tokenane
Feature activation+0.000
z
Tokenz
Feature activation+0.000
@
Token@
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+1.246
H
TokenH
Feature activation+0.000
ollywood
Tokenollywood
Feature activation+0.000
's
Token's
Feature activation+0.000
highest
Token highest
Feature activation+0.000
profile
Token profile
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
izers
Tokenizers
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ŀ
TokenĿ
Feature activation+0.000
âĢĵ
Token âĢĵ
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+1.404
Due
TokenDue
Feature activation+0.000
to
Token to
Feature activation+0.000
boring
Token boring
Feature activation+0.000
circumstances
Token circumstances
Feature activation+0.000
beyond
Token beyond
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
Like
Token Like
Feature activation+0.000
Loading
Token Loading
Feature activation+0.000
...
Token...
Feature activation+0.000
Related
Token Related
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+1.291
5
Token5
Feature activation+0.000
.
Token.
Feature activation+0.000
0
Token0
Feature activation+0.000
âĺħ
Token âĺħ
Feature activation+0.000
âĺħâĺħ
Tokenâĺħâĺħ
Feature activation+0.000

INTERVAL 0.807 - 1.210
CONTAINS 0.001%

<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
the
Token the
Feature activation+0.000
powerful
Token powerful
Feature activation+0.000
incentive
Token incentive
Feature activation+0.000
of
Token of
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+1.063
Police
TokenPolice
Feature activation+0.000
snap
Token snap
Feature activation+0.000
up
Token up
Feature activation+0.000
mud
Token mud
Feature activation+0.000
crab
Token crab
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
ofi
Tokenofi
Feature activation+0.000
was
Token was
Feature activation+0.000
spectacular
Token spectacular
Feature activation+0.000
and
Token and
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+1.188
Black
TokenBlack
Feature activation+0.000
Girls
Token Girls
Feature activation+0.000
Rock
Token Rock
Feature activation+0.000
!
Token!
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
sc
Tokensc
Feature activation+0.000
ast
Tokenast
Feature activation+0.000
unit
Token unit
Feature activation+0.000
:
Token:
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+1.066
Wars
TokenWars
Feature activation+0.000
were
Token were
Feature activation+0.000
fought
Token fought
Feature activation+0.000
to
Token to
Feature activation+0.000
impose
Token impose
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
:
Token:
Feature activation+0.000
1993
Token 1993
Feature activation+0.000
âĢĵ
TokenâĢĵ
Feature activation+0.000
1995
Token1995
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.952
Rel
TokenRel
Feature activation+0.000
igious
Tokenigious
Feature activation+0.000
leaders
Token leaders
Feature activation+0.000
,
Token,
Feature activation+0.000
including
Token including
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
the
Token the
Feature activation+0.000
number
Token number
Feature activation+0.000
of
Token of
Feature activation+0.000
highly
Token highly
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.841
MS
TokenMS
Feature activation+0.000
NBC
TokenNBC
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
s
Tokens
Feature activation+0.000

INTERVAL 0.403 - 0.807
CONTAINS 0.001%

<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
for
Token for
Feature activation+0.000
the
Token the
Feature activation+0.000
future
Token future
Feature activation+0.000
."
Token."
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.788
DON
TokenDON
Feature activation+0.000
ALD
TokenALD
Feature activation+0.000
Trump
Token Trump
Feature activation+0.000
wasn
Token wasn
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
@
Token@
Feature activation+0.000
gmail
Tokengmail
Feature activation+0.000
.
Token.
Feature activation+0.000
com
Tokencom
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.671
Keith
TokenKeith
Feature activation+0.000
Law
Token Law
Feature activation+0.000
gets
Token gets
Feature activation+0.000
Twitter
Token Twitter
Feature activation+0.000
suspension
Token suspension
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
Senator
Token Senator
Feature activation+0.000
Ted
Token Ted
Feature activation+0.000
Cruz
Token Cruz
Feature activation+0.000
.
Token.
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.575
Anthony
TokenAnthony
Feature activation+0.000
Log
Token Log
Feature activation+0.000
istics
Tokenistics
Feature activation+0.000
for
Token for
Feature activation+0.000
Men
Token Men
Feature activation+0.000
pushing
Token pushing
Feature activation+0.000
for
Token for
Feature activation+0.000
similar
Token similar
Feature activation+0.000
outcomes
Token outcomes
Feature activation+0.000
for
Token for
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.413
Gl
TokenGl
Feature activation+0.000
obe
Tokenobe
Feature activation+0.000
-
Token-
Feature activation+0.000
t
Tokent
Feature activation+0.000
rot
Tokenrot
Feature activation+0.000
afternoon
Token afternoon
Feature activation+0.000
before
Token before
Feature activation+0.000
she
Token she
Feature activation+0.000
disappeared
Token disappeared
Feature activation+0.000
.
Token.
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.444
We
TokenWe
Feature activation+0.000
suggest
Token suggest
Feature activation+0.000
that
Token that
Feature activation+0.000
signal
Token signal
Feature activation+0.000
convergence
Token convergence
Feature activation+0.000

INTERVAL 0.000 - 0.403
CONTAINS 99.993%

wallpaper
Token wallpaper
Feature activation+0.000
versions
Token versions
Feature activation+0.000
of
Token of
Feature activation+0.000
Thor
Token Thor
Feature activation+0.000
's
Token's
Feature activation+0.000
art
Token art
Feature activation+0.000
seen
Token seen
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
campaign
Token campaign
Feature activation+0.000
video
Token video
Feature activation+0.000
They
Token They
Feature activation+0.000
are
Token are
Feature activation+0.000
more
Token more
Feature activation+0.000
of
Token of
Feature activation+0.000
a
Token a
Feature activation+0.000
âĢ
Token âĢ
Feature activation+0.000
ľ
Tokenľ
Feature activation+0.000
day
Tokenday
Feature activation+0.000
2
Token 2
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ŀ
TokenĿ
Feature activation+0.000
were
Token were
Feature activation+0.000
able
Token able
Feature activation+0.000
to
Token to
Feature activation+0.000
prevent
Token prevent
Feature activation+0.000
a
Token a
Feature activation+0.000
North
Token North
Feature activation+0.000
Korean
Token Korean
Feature activation+0.000
division
Token division
Feature activation+0.000
from
Token from
Feature activation+0.000
capturing
Token capturing
Feature activation+0.000
the
Token the
Feature activation+0.000
plet
Tokenplet
Feature activation+0.000
of
Token of
Feature activation+0.000
water
Token water
Feature activation+0.000
âĢĶ
Token âĢĶ
Feature activation+0.000
could
Token could
Feature activation+0.000
oscill
Token oscill
Feature activation+0.000
ate
Tokenate
Feature activation+0.000
back
Token back
Feature activation+0.000
and
Token and
Feature activation+0.000
forth
Token forth
Feature activation+0.000
fast
Token fast
Feature activation+0.000
wrote
Token wrote
Feature activation+0.000
Jason
Token Jason
Feature activation+0.000
Dor
Token Dor
Feature activation+0.000
rier
Tokenrier
Feature activation+0.000
with
Token with
Feature activation+0.000
the
Token the
Feature activation+0.000
Sing
Token Sing
Feature activation+0.000
ularity
Tokenularity
Feature activation+0.000
Hub
Token Hub
Feature activation+0.000
.
Token.
Feature activation+0.000
âĢ
Token âĢ
Feature activation+0.000

BOTTOM ACTIVATIONS
MIN = 0.000

regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
anking
Tokenanking
Feature activation+0.000
the
Token the
Feature activation+0.000
sc
Token sc
Feature activation+0.000
iss
Tokeniss
Feature activation+0.000
ile
Tokenile
Feature activation+0.000
bonds
Token bonds
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
anking
Tokenanking
Feature activation+0.000
the
Token the
Feature activation+0.000
sc
Token sc
Feature activation+0.000
iss
Tokeniss
Feature activation+0.000
ile
Tokenile
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
anking
Tokenanking
Feature activation+0.000
the
Token the
Feature activation+0.000
sc
Token sc
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
anking
Tokenanking
Feature activation+0.000
the
Token the
Feature activation+0.000
sc
Token sc
Feature activation+0.000
iss
Tokeniss
Feature activation+0.000
ile
Tokenile
Feature activation+0.000
(
Token (
Feature activation+0.000
Michael
TokenMichael
Feature activation+0.000
C
Token C
Feature activation+0.000
aine
Tokenaine
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
ile
Tokenile
Feature activation+0.000
(
Token (
Feature activation+0.000
Michael
TokenMichael
Feature activation+0.000
C
Token C
Feature activation+0.000
aine
Tokenaine
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
(
Token (
Feature activation+0.000
Michael
TokenMichael
Feature activation+0.000
C
Token C
Feature activation+0.000
aine
Tokenaine
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Michael
TokenMichael
Feature activation+0.000
C
Token C
Feature activation+0.000
aine
Tokenaine
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
anking
Tokenanking
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
anking
Tokenanking
Feature activation+0.000
the
Token the
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
C
Token C
Feature activation+0.000
aine
Tokenaine
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
aine
Tokenaine
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000

Top feature 6 in H0.11: (feature 4960

TOP ACTIVATIONS
MAX = 2.817

<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
thing
Token thing
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
s
Tokens
Feature activation+0.000
not
Token not
Feature activation+1.261
where
Token where
Feature activation+0.000
it
Token it
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
s
Tokens
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Chris
TokenChris
Feature activation+0.000
Matthews
Token Matthews
Feature activation+0.000
did
Token did
Feature activation+0.000
not
Token not
Feature activation+1.242
return
Token return
Feature activation+0.000
calls
Token calls
Feature activation+0.000
for
Token for
Feature activation+0.000
comment
Token comment
Feature activation+0.000
.
Token.
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
It
Token It
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
s
Tokens
Feature activation+0.000
not
Token not
Feature activation+1.225
what
Token what
Feature activation+0.000
I
Token I
Feature activation+0.000
meant
Token meant
Feature activation+0.000
.
Token.
Feature activation+0.000
I
Token I
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
it
Token it
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
s
Tokens
Feature activation+0.000
not
Token not
Feature activation+1.169
true
Token true
Feature activation+0.000
.
Token.
Feature activation+0.000
According
Token According
Feature activation+0.000
to
Token to
Feature activation+0.000
the
Token the
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
company
Token company
Feature activation+0.000
did
Token did
Feature activation+0.000
not
Token not
Feature activation+1.107
immediately
Token immediately
Feature activation+0.000
return
Token return
Feature activation+0.000
a
Token a
Feature activation+0.000
request
Token request
Feature activation+0.000
for
Token for
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
complicated
Token complicated
Feature activation+0.000
and
Token and
Feature activation+0.000
nuanced
Token nuanced
Feature activation+0.000
tale
Token tale
Feature activation+0.000
not
Token not
Feature activation+1.097
merely
Token merely
Feature activation+0.000
of
Token of
Feature activation+0.000
love
Token love
Feature activation+0.000
gone
Token gone
Feature activation+0.000
bad
Token bad
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
the
Token the
Feature activation+0.000
United
Token United
Feature activation+0.000
States
Token States
Feature activation+0.000
,
Token,
Feature activation+0.000
not
Token not
Feature activation+1.063
all
Token all
Feature activation+0.000
of
Token of
Feature activation+0.000
its
Token its
Feature activation+0.000
parts
Token parts
Feature activation+0.000
would
Token would
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
.
Token.
Feature activation+0.000
Corpor
Token Corpor
Feature activation+0.000
ations
Tokenations
Feature activation+0.000
are
Token are
Feature activation+0.000
not
Token not
Feature activation+1.000
to
Token to
Feature activation+0.000
blame
Token blame
Feature activation+0.000
.
Token.
Feature activation+0.000
Banks
Token Banks
Feature activation+0.000
are
Token are
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
âĢ
Token âĢ
Feature activation+0.000
ľ
Tokenľ
Feature activation+0.000
We
TokenWe
Feature activation+0.000
could
Token could
Feature activation+0.000
not
Token not
Feature activation+0.997
pay
Token pay
Feature activation+0.000
technicians
Token technicians
Feature activation+0.000
and
Token and
Feature activation+0.000
suppliers
Token suppliers
Feature activation+0.000
of
Token of
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
credibility
Token credibility
Feature activation+0.000
.
Token.
Feature activation+0.000
It
TokenIt
Feature activation+0.000
's
Token's
Feature activation+0.000
not
Token not
Feature activation+0.984
just
Token just
Feature activation+0.000
that
Token that
Feature activation+0.000
,
Token,
Feature activation+0.000
it
Token it
Feature activation+0.000
's
Token's
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
election
Token election
Feature activation+0.000
,
Token,
Feature activation+0.000
warning
Token warning
Feature activation+0.000
people
Token people
Feature activation+0.000
not
Token not
Feature activation+0.980
to
Token to
Feature activation+0.000
"
Token "
Feature activation+0.000
und
Tokenund
Feature activation+0.000
erest
Tokenerest
Feature activation+0.000
imate
Tokenimate
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
however
Token however
Feature activation+0.000
,
Token,
Feature activation+0.000
this
Token this
Feature activation+0.000
did
Token did
Feature activation+0.000
not
Token not
Feature activation+0.971
happen
Token happen
Feature activation+0.000
as
Token as
Feature activation+0.000
the
Token the
Feature activation+0.000
militia
Token militia
Feature activation+0.000
failed
Token failed
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
Going
Token Going
Feature activation+0.000
it
Token it
Feature activation+0.000
alone
Token alone
Feature activation+0.000
is
Token is
Feature activation+0.000
not
Token not
Feature activation+0.956
the
Token the
Feature activation+0.000
end
Token end
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
world
Token world
Feature activation+0.000
,
Token,
Feature activation+0.000
why
Token why
Feature activation+0.000
not
Token not
Feature activation+2.040
?
Token?
Feature activation+0.000
Why
Token Why
Feature activation+0.000
not
Token not
Feature activation+0.920
marry
Token marry
Feature activation+0.000
a
Token a
Feature activation+0.000
giant
Token giant
Feature activation+0.000
blood
Token blood
Feature activation+0.000
-
Token-
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
time
Token time
Feature activation+0.000
,
Token,
Feature activation+0.000
authorities
Token authorities
Feature activation+0.000
do
Token do
Feature activation+0.000
not
Token not
Feature activation+0.898
believe
Token believe
Feature activation+0.000
someone
Token someone
Feature activation+0.000
deliberately
Token deliberately
Feature activation+0.000
poisoned
Token poisoned
Feature activation+0.000
the
Token the
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
parents
Token parents
Feature activation+0.000
.
Token.
Feature activation+0.000
It
Token It
Feature activation+0.000
does
Token does
Feature activation+0.000
not
Token not
Feature activation+0.895
recognise
Token recognise
Feature activation+0.000
social
Token social
Feature activation+0.000
status
Token status
Feature activation+0.000
.
Token.
Feature activation+0.000
The
Token The
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
Additionally
Token Additionally
Feature activation+0.000
,
Token,
Feature activation+0.000
it
Token it
Feature activation+0.000
was
Token was
Feature activation+0.000
not
Token not
Feature activation+0.894
hard
Token hard
Feature activation+0.000
to
Token to
Feature activation+0.000
see
Token see
Feature activation+0.000
during
Token during
Feature activation+0.000
this
Token this
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
the
Token the
Feature activation+0.000
incident
Token incident
Feature activation+0.000
,
Token,
Feature activation+0.000
but
Token but
Feature activation+0.000
not
Token not
Feature activation+0.887
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
investigation
Token investigation
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
is
Token is
Feature activation+0.000
on
Token on
Feature activation+0.000
capital
Token capital
Feature activation+0.000
,
Token,
Feature activation+0.000
not
Token not
Feature activation+0.876
innovation
Token innovation
Feature activation+0.000
.
Token.
Feature activation+0.000
The
Token The
Feature activation+0.000
markets
Token markets
Feature activation+0.000
are
Token are
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
hoping
Token hoping
Feature activation+0.000
for
Token for
Feature activation+0.000
an
Token an
Feature activation+0.000
edge
Token edge
Feature activation+0.000
not
Token not
Feature activation+0.858
just
Token just
Feature activation+0.000
with
Token with
Feature activation+0.000
the
Token the
Feature activation+0.000
Spotify
Token Spotify
Feature activation+0.000
association
Token association
Feature activation+0.000

Top DFA by src position
MAX = 6.452

<|endoftext|>
Token<|endoftext|>
Feature activation+0.889
Top resid features:
thing
Token thing
Feature activation+0.032
Top resid features:
âĢ
TokenâĢ
Feature activation-0.171
Top resid features:
Ļ
TokenĻ
Feature activation-0.028
Top resid features:
s
Tokens
Feature activation+0.327
Top resid features:
not
Token not
Feature activation+6.452
Top resid features:
where
Token where
Feature activation+0.000
Top resid features:
it
Token it
Feature activation+0.000
Top resid features:
âĢ
TokenâĢ
Feature activation+0.000
Top resid features:
Ļ
TokenĻ
Feature activation+0.000
Top resid features:
s
Tokens
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.852
Top resid features:
Ċ
TokenĊ
Feature activation+0.016
Top resid features:
Chris
TokenChris
Feature activation-0.080
Top resid features:
Matthews
Token Matthews
Feature activation-0.034
Top resid features:
did
Token did
Feature activation+1.816
Top resid features:
not
Token not
Feature activation+4.911
Top resid features:
return
Token return
Feature activation+0.000
Top resid features:
calls
Token calls
Feature activation+0.000
Top resid features:
for
Token for
Feature activation+0.000
Top resid features:
comment
Token comment
Feature activation+0.000
Top resid features:
.
Token.
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.849
Top resid features:
It
Token It
Feature activation+0.293
Top resid features:
âĢ
TokenâĢ
Feature activation-0.198
Top resid features:
Ļ
TokenĻ
Feature activation-0.045
Top resid features:
s
Tokens
Feature activation+0.283
Top resid features:
not
Token not
Feature activation+6.283
Top resid features:
what
Token what
Feature activation+0.000
Top resid features:
I
Token I
Feature activation+0.000
Top resid features:
meant
Token meant
Feature activation+0.000
Top resid features:
.
Token.
Feature activation+0.000
Top resid features:
I
Token I
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.891
Top resid features:
it
Token it
Feature activation+0.240
Top resid features:
âĢ
TokenâĢ
Feature activation-0.192
Top resid features:
Ļ
TokenĻ
Feature activation-0.049
Top resid features:
s
Tokens
Feature activation+0.260
Top resid features:
not
Token not
Feature activation+6.258
Top resid features:
true
Token true
Feature activation+0.000
Top resid features:
.
Token.
Feature activation+0.000
Top resid features:
According
Token According
Feature activation+0.000
Top resid features:
to
Token to
Feature activation+0.000
Top resid features:
the
Token the
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.725
Top resid features:
Ċ
TokenĊ
Feature activation-0.040
Top resid features:
The
TokenThe
Feature activation+0.243
Top resid features:
company
Token company
Feature activation-0.081
Top resid features:
did
Token did
Feature activation+1.747
Top resid features:
not
Token not
Feature activation+4.753
Top resid features:
immediately
Token immediately
Feature activation+0.000
Top resid features:
return
Token return
Feature activation+0.000
Top resid features:
a
Token a
Feature activation+0.000
Top resid features:
request
Token request
Feature activation+0.000
Top resid features:
for
Token for
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.823
Top resid features:
complicated
Token complicated
Feature activation+0.033
Top resid features:
and
Token and
Feature activation+0.320
Top resid features:
nuanced
Token nuanced
Feature activation-0.065
Top resid features:
tale
Token tale
Feature activation+0.031
Top resid features:
not
Token not
Feature activation+6.195
Top resid features:
merely
Token merely
Feature activation+0.000
Top resid features:
of
Token of
Feature activation+0.000
Top resid features:
love
Token love
Feature activation+0.000
Top resid features:
gone
Token gone
Feature activation+0.000
Top resid features:
bad
Token bad
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.811
Top resid features:
the
Token the
Feature activation+0.306
Top resid features:
United
Token United
Feature activation-0.052
Top resid features:
States
Token States
Feature activation-0.072
Top resid features:
,
Token,
Feature activation+0.234
Top resid features:
not
Token not
Feature activation+6.075
Top resid features:
all
Token all
Feature activation+0.000
Top resid features:
of
Token of
Feature activation+0.000
Top resid features:
its
Token its
Feature activation+0.000
Top resid features:
parts
Token parts
Feature activation+0.000
Top resid features:
would
Token would
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.715
Top resid features:
.
Token.
Feature activation+0.015
Top resid features:
Corpor
Token Corpor
Feature activation-0.080
Top resid features:
ations
Tokenations
Feature activation-0.056
Top resid features:
are
Token are
Feature activation+0.744
Top resid features:
not
Token not
Feature activation+5.902
Top resid features:
to
Token to
Feature activation+0.000
Top resid features:
blame
Token blame
Feature activation+0.000
Top resid features:
.
Token.
Feature activation+0.000
Top resid features:
Banks
Token Banks
Feature activation+0.000
Top resid features:
are
Token are
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.817
Top resid features:
âĢ
Token âĢ
Feature activation-0.155
Top resid features:
ľ
Tokenľ
Feature activation+0.068
Top resid features:
We
TokenWe
Feature activation-0.040
Top resid features:
could
Token could
Feature activation+1.561
Top resid features:
not
Token not
Feature activation+4.986
Top resid features:
pay
Token pay
Feature activation+0.000
Top resid features:
technicians
Token technicians
Feature activation+0.000
Top resid features:
and
Token and
Feature activation+0.000
Top resid features:
suppliers
Token suppliers
Feature activation+0.000
Top resid features:
of
Token of
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.679
Top resid features:
credibility
Token credibility
Feature activation+0.048
Top resid features:
.
Token.
Feature activation+0.006
Top resid features:
It
TokenIt
Feature activation+0.174
Top resid features:
's
Token's
Feature activation+0.296
Top resid features:
not
Token not
Feature activation+6.020
Top resid features:
just
Token just
Feature activation+0.000
Top resid features:
that
Token that
Feature activation+0.000
Top resid features:
,
Token,
Feature activation+0.000
Top resid features:
it
Token it
Feature activation+0.000
Top resid features:
's
Token's
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.778
Top resid features:
election
Token election
Feature activation+0.054
Top resid features:
,
Token,
Feature activation+0.252
Top resid features:
warning
Token warning
Feature activation-0.163
Top resid features:
people
Token people
Feature activation+0.048
Top resid features:
not
Token not
Feature activation+6.251
Top resid features:
to
Token to
Feature activation+0.000
Top resid features:
"
Token "
Feature activation+0.000
Top resid features:
und
Tokenund
Feature activation+0.000
Top resid features:
erest
Tokenerest
Feature activation+0.000
Top resid features:
imate
Tokenimate
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.734
Top resid features:
however
Token however
Feature activation-0.118
Top resid features:
,
Token,
Feature activation+0.148
Top resid features:
this
Token this
Feature activation+0.148
Top resid features:
did
Token did
Feature activation+1.685
Top resid features:
not
Token not
Feature activation+4.613
Top resid features:
happen
Token happen
Feature activation+0.000
Top resid features:
as
Token as
Feature activation+0.000
Top resid features:
the
Token the
Feature activation+0.000
Top resid features:
militia
Token militia
Feature activation+0.000
Top resid features:
failed
Token failed
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.677
Top resid features:
Going
Token Going
Feature activation+0.011
Top resid features:
it
Token it
Feature activation+0.153
Top resid features:
alone
Token alone
Feature activation-0.059
Top resid features:
is
Token is
Feature activation+0.527
Top resid features:
not
Token not
Feature activation+5.886
Top resid features:
the
Token the
Feature activation+0.000
Top resid features:
end
Token end
Feature activation+0.000
Top resid features:
of
Token of
Feature activation+0.000
Top resid features:
the
Token the
Feature activation+0.000
Top resid features:
world
Token world
Feature activation+0.000
Top resid features:
,
Token,
Feature activation+0.214
Top resid features:
why
Token why
Feature activation+0.078
Top resid features:
not
Token not
Feature activation+2.333
Top resid features:
?
Token?
Feature activation+0.031
Top resid features:
Why
Token Why
Feature activation+1.078
Top resid features:
not
Token not
Feature activation+3.006
Top resid features:
marry
Token marry
Feature activation+0.000
Top resid features:
a
Token a
Feature activation+0.000
Top resid features:
giant
Token giant
Feature activation+0.000
Top resid features:
blood
Token blood
Feature activation+0.000
Top resid features:
-
Token-
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.797
Top resid features:
time
Token time
Feature activation+0.039
Top resid features:
,
Token,
Feature activation+0.195
Top resid features:
authorities
Token authorities
Feature activation-0.162
Top resid features:
do
Token do
Feature activation+0.811
Top resid features:
not
Token not
Feature activation+5.458
Top resid features:
believe
Token believe
Feature activation+0.000
Top resid features:
someone
Token someone
Feature activation+0.000
Top resid features:
deliberately
Token deliberately
Feature activation+0.000
Top resid features:
poisoned
Token poisoned
Feature activation+0.000
Top resid features:
the
Token the
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.640
Top resid features:
parents
Token parents
Feature activation-0.027
Top resid features:
.
Token.
Feature activation+0.022
Top resid features:
It
Token It
Feature activation+0.297
Top resid features:
does
Token does
Feature activation+0.898
Top resid features:
not
Token not
Feature activation+5.304
Top resid features:
recognise
Token recognise
Feature activation+0.000
Top resid features:
social
Token social
Feature activation+0.000
Top resid features:
status
Token status
Feature activation+0.000
Top resid features:
.
Token.
Feature activation+0.000
Top resid features:
The
Token The
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.723
Top resid features:
Additionally
Token Additionally
Feature activation-0.101
Top resid features:
,
Token,
Feature activation+0.077
Top resid features:
it
Token it
Feature activation+0.192
Top resid features:
was
Token was
Feature activation+0.502
Top resid features:
not
Token not
Feature activation+5.741
Top resid features:
hard
Token hard
Feature activation+0.000
Top resid features:
to
Token to
Feature activation+0.000
Top resid features:
see
Token see
Feature activation+0.000
Top resid features:
during
Token during
Feature activation+0.000
Top resid features:
this
Token this
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.752
Top resid features:
the
Token the
Feature activation+0.288
Top resid features:
incident
Token incident
Feature activation-0.087
Top resid features:
,
Token,
Feature activation+0.184
Top resid features:
but
Token but
Feature activation+0.021
Top resid features:
not
Token not
Feature activation+5.967
Top resid features:
of
Token of
Feature activation+0.000
Top resid features:
the
Token the
Feature activation+0.000
Top resid features:
investigation
Token investigation
Feature activation+0.000
Top resid features:
.
Token.
Feature activation+0.000
Top resid features:
Ċ
TokenĊ
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.718
Top resid features:
is
Token is
Feature activation+0.255
Top resid features:
on
Token on
Feature activation+0.047
Top resid features:
capital
Token capital
Feature activation-0.106
Top resid features:
,
Token,
Feature activation+0.194
Top resid features:
not
Token not
Feature activation+6.007
Top resid features:
innovation
Token innovation
Feature activation+0.000
Top resid features:
.
Token.
Feature activation+0.000
Top resid features:
The
Token The
Feature activation+0.000
Top resid features:
markets
Token markets
Feature activation+0.000
Top resid features:
are
Token are
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.828
Top resid features:
hoping
Token hoping
Feature activation-0.112
Top resid features:
for
Token for
Feature activation+0.164
Top resid features:
an
Token an
Feature activation+0.131
Top resid features:
edge
Token edge
Feature activation-0.071
Top resid features:
not
Token not
Feature activation+6.157
Top resid features:
just
Token just
Feature activation+0.000
Top resid features:
with
Token with
Feature activation+0.000
Top resid features:
the
Token the
Feature activation+0.000
Top resid features:
Spotify
Token Spotify
Feature activation+0.000
Top resid features:
association
Token association
Feature activation+0.000
Top resid features:

Decoder Weights Distribution

Head 0: 0.05

Head 1: 0.12

Head 2: 0.06

Head 3: 0.11

Head 4: 0.05

Head 5: 0.07

Head 6: 0.07

Head 7: 0.07

Head 8: 0.06

Head 9: 0.14

Head 10: 0.07

Head 11: 0.14

Positive logits

etheless3.70

NetMessage3.45

proble3.44

confir3.40

mathemat3.12

nutrit3.07

VIDIA3.03

issan2.91

andem2.82

lihood2.79

��2.74

aukee2.71

destro2.70

terday2.65

icably2.65

cffff2.64

ibilities2.63

suspic2.63

appre2.63

unden2.62

Negative logits

FactoryReloaded-2.60

town-2.58

ItemLevel-2.43

lace-2.40

CLSID-2.35

Tokens-2.24

front-2.23

arsity-2.22

"}],"-2.21

Discussion-2.15

eers-2.06

Intern-2.05

girls-2.05

ーティ-2.04

breakers-2.03

eering-2.03

Conversation-2.01

girl-2.00

thesis-2.00

itect-1.99

INTERVAL 2.535 - 2.817
CONTAINS 0.002%

INTERVAL 2.254 - 2.535
CONTAINS 0.002%

INTERVAL 1.972 - 2.254
CONTAINS 0.000%

INTERVAL 1.690 - 1.972
CONTAINS 0.001%

INTERVAL 1.409 - 1.690
CONTAINS 0.001%

INTERVAL 1.127 - 1.409
CONTAINS 0.003%

<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
it
Token it
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
s
Tokens
Feature activation+0.000
not
Token not
Feature activation+1.169
true
Token true
Feature activation+0.000
.
Token.
Feature activation+0.000
According
Token According
Feature activation+0.000
to
Token to
Feature activation+0.000
the
Token the
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
thing
Token thing
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
s
Tokens
Feature activation+0.000
not
Token not
Feature activation+1.261
where
Token where
Feature activation+0.000
it
Token it
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
s
Tokens
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Chris
TokenChris
Feature activation+0.000
Matthews
Token Matthews
Feature activation+0.000
did
Token did
Feature activation+0.000
not
Token not
Feature activation+1.242
return
Token return
Feature activation+0.000
calls
Token calls
Feature activation+0.000
for
Token for
Feature activation+0.000
comment
Token comment
Feature activation+0.000
.
Token.
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
It
Token It
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
s
Tokens
Feature activation+0.000
not
Token not
Feature activation+1.225
what
Token what
Feature activation+0.000
I
Token I
Feature activation+0.000
meant
Token meant
Feature activation+0.000
.
Token.
Feature activation+0.000
I
Token I
Feature activation+0.000

INTERVAL 0.845 - 1.127
CONTAINS 0.002%

<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
the
Token the
Feature activation+0.000
United
Token United
Feature activation+0.000
States
Token States
Feature activation+0.000
,
Token,
Feature activation+0.000
not
Token not
Feature activation+1.063
all
Token all
Feature activation+0.000
of
Token of
Feature activation+0.000
its
Token its
Feature activation+0.000
parts
Token parts
Feature activation+0.000
would
Token would
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
however
Token however
Feature activation+0.000
,
Token,
Feature activation+0.000
this
Token this
Feature activation+0.000
did
Token did
Feature activation+0.000
not
Token not
Feature activation+0.971
happen
Token happen
Feature activation+0.000
as
Token as
Feature activation+0.000
the
Token the
Feature activation+0.000
militia
Token militia
Feature activation+0.000
failed
Token failed
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
.
Token.
Feature activation+0.000
Corpor
Token Corpor
Feature activation+0.000
ations
Tokenations
Feature activation+0.000
are
Token are
Feature activation+0.000
not
Token not
Feature activation+1.000
to
Token to
Feature activation+0.000
blame
Token blame
Feature activation+0.000
.
Token.
Feature activation+0.000
Banks
Token Banks
Feature activation+0.000
are
Token are
Feature activation+0.000
standard
Tokenstandard
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ŀ
TokenĿ
Feature activation+0.000
mechanisms
Token mechanisms
Feature activation+0.000
should
Token should
Feature activation+0.000
not
Token not
Feature activation+0.854
be
Token be
Feature activation+0.000
disreg
Token disreg
Feature activation+0.000
arded
Tokenarded
Feature activation+0.000
.
Token.
Feature activation+0.000
For
Token For
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
time
Token time
Feature activation+0.000
,
Token,
Feature activation+0.000
authorities
Token authorities
Feature activation+0.000
do
Token do
Feature activation+0.000
not
Token not
Feature activation+0.898
believe
Token believe
Feature activation+0.000
someone
Token someone
Feature activation+0.000
deliberately
Token deliberately
Feature activation+0.000
poisoned
Token poisoned
Feature activation+0.000
the
Token the
Feature activation+0.000

INTERVAL 0.563 - 0.845
CONTAINS 0.001%

App
Token App
Feature activation+0.000
:
Token:
Feature activation+0.000
Your
Token Your
Feature activation+0.000
friend
Token friend
Feature activation+0.000
will
Token will
Feature activation+0.000
not
Token not
Feature activation+0.679
get
Token get
Feature activation+0.000
their
Token their
Feature activation+0.000
Attend
Token Attend
Feature activation+0.000
ance
Tokenance
Feature activation+0.000
app
Token app
Feature activation+0.000
video
Token video
Feature activation+0.000
saying
Token saying
Feature activation+0.000
the
Token the
Feature activation+0.000
man
Token man
Feature activation+0.000
was
Token was
Feature activation+0.000
not
Token not
Feature activation+0.623
armed
Token armed
Feature activation+0.000
,
Token,
Feature activation+0.000
but
Token but
Feature activation+0.000
police
Token police
Feature activation+0.000
say
Token say
Feature activation+0.000
rell
Tokenrell
Feature activation+0.000
said
Token said
Feature activation+0.000
the
Token the
Feature activation+0.000
patient
Token patient
Feature activation+0.000
had
Token had
Feature activation+0.000
not
Token not
Feature activation+0.707
been
Token been
Feature activation+0.000
affected
Token affected
Feature activation+0.000
the
Token the
Feature activation+0.000
way
Token way
Feature activation+0.000
consortium
Token consortium
Feature activation+0.000
about
Token about
Feature activation+0.000
sincerity
Token sincerity
Feature activation+0.000
if
Token if
Feature activation+0.000
it
Token it
Feature activation+0.000
leads
Token leads
Feature activation+0.000
not
Token not
Feature activation+0.638
to
Token to
Feature activation+0.000
understanding
Token understanding
Feature activation+0.000
but
Token but
Feature activation+0.000
to
Token to
Feature activation+0.000
myst
Token myst
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
but
Token but
Feature activation+0.000
the
Token the
Feature activation+0.000
marriage
Token marriage
Feature activation+0.000
was
Token was
Feature activation+0.000
not
Token not
Feature activation+0.807
happy
Token happy
Feature activation+0.000
.
Token.
Feature activation+0.000
In
Token In
Feature activation+0.000
15
Token 15
Feature activation+0.000
32
Token32
Feature activation+0.000

INTERVAL 0.282 - 0.563
CONTAINS 0.002%

be
Token be
Feature activation+0.000
discouraged
Token discouraged
Feature activation+0.000
if
Token if
Feature activation+0.000
they
Token they
Feature activation+0.000
're
Token're
Feature activation+0.000
not
Token not
Feature activation+0.389
successful
Token successful
Feature activation+0.000
immediately
Token immediately
Feature activation+0.000
.
Token.
Feature activation+0.000
"
Token "
Feature activation+0.000
It
TokenIt
Feature activation+0.000
se
Tokense
Feature activation+0.000
holes
Tokenholes
Feature activation+0.000
.
Token.
Feature activation+0.000
That
Token That
Feature activation+0.000
is
Token is
Feature activation+0.000
not
Token not
Feature activation+0.312
a
Token a
Feature activation+0.000
miracle
Token miracle
Feature activation+0.000
.
Token.
Feature activation+0.000
It
Token It
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
take
Token take
Feature activation+0.000
action
Token action
Feature activation+0.000
.
Token.
Feature activation+0.000
They
Token They
Feature activation+0.000
do
Token do
Feature activation+0.000
not
Token not
Feature activation+0.347
want
Token want
Feature activation+0.000
to
Token to
Feature activation+0.000
feel
Token feel
Feature activation+0.000
complicit
Token complicit
Feature activation+0.000
in
Token in
Feature activation+0.000
court
Token court
Feature activation+0.000
said
Token said
Feature activation+0.000
the
Token the
Feature activation+0.000
shortfall
Token shortfall
Feature activation+0.000
was
Token was
Feature activation+0.000
not
Token not
Feature activation+0.525
due
Token due
Feature activation+0.000
to
Token to
Feature activation+0.000
a
Token a
Feature activation+0.000
long
Token long
Feature activation+0.000
-
Token-
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
First
Token First
Feature activation+0.000
Amendment
Token Amendment
Feature activation+0.000
would
Token would
Feature activation+0.000
not
Token not
Feature activation+0.344
be
Token be
Feature activation+0.000
opposed
Token opposed
Feature activation+0.000
."
Token."
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000

INTERVAL 0.000 - 0.282
CONTAINS 99.986%

)
Token)
Feature activation+0.000
sophisticated
Token sophisticated
Feature activation+0.000
counterfe
Token counterfe
Feature activation+0.000
iting
Tokeniting
Feature activation+0.000
operations
Token operations
Feature activation+0.000
operating
Token operating
Feature activation+0.000
out
Token out
Feature activation+0.000
of
Token of
Feature activation+0.000
Lima
Token Lima
Feature activation+0.000
,
Token,
Feature activation+0.000
Peru
Token Peru
Feature activation+0.000
aut
Tokenaut
Feature activation+0.000
ner
Tokenner
Feature activation+0.000
,
Token,
Feature activation+0.000
the
Token the
Feature activation+0.000
son
Token son
Feature activation+0.000
of
Token of
Feature activation+0.000
an
Token an
Feature activation+0.000
academic
Token academic
Feature activation+0.000
and
Token and
Feature activation+0.000
an
Token an
Feature activation+0.000
artist
Token artist
Feature activation+0.000
the
Token the
Feature activation+0.000
world
Token world
Feature activation+0.000
âĢĵ
Token âĢĵ
Feature activation+0.000
the
Token the
Feature activation+0.000
world
Token world
Feature activation+0.000
where
Token where
Feature activation+0.000
I
Token I
Feature activation+0.000
already
Token already
Feature activation+0.000
live
Token live
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
)
Token)
Feature activation+0.000
and
Token and
Feature activation+0.000
1
Token 1
Feature activation+0.000
,
Token,
Feature activation+0.000
650
Token650
Feature activation+0.000
-
Token-
Feature activation+0.000
yard
Tokenyard
Feature activation+0.000
fre
Token fre
Feature activation+0.000
estyle
Tokenestyle
Feature activation+0.000
(
Token (
Feature activation+0.000
Jessica
TokenJessica
Feature activation+0.000
you
Token you
Feature activation+0.000
learn
Token learn
Feature activation+0.000
about
Token about
Feature activation+0.000
what
Token what
Feature activation+0.000
is
Token is
Feature activation+0.000
your
Token your
Feature activation+0.000
passion
Token passion
Feature activation+0.000
,
Token,
Feature activation+0.000
but
Token but
Feature activation+0.000
also
Token also
Feature activation+0.000
make
Token make
Feature activation+0.000

BOTTOM ACTIVATIONS
MIN = 0.000

regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
anking
Tokenanking
Feature activation+0.000
the
Token the
Feature activation+0.000
sc
Token sc
Feature activation+0.000
iss
Tokeniss
Feature activation+0.000
ile
Tokenile
Feature activation+0.000
bonds
Token bonds
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
anking
Tokenanking
Feature activation+0.000
the
Token the
Feature activation+0.000
sc
Token sc
Feature activation+0.000
iss
Tokeniss
Feature activation+0.000
ile
Tokenile
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
anking
Tokenanking
Feature activation+0.000
the
Token the
Feature activation+0.000
sc
Token sc
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
anking
Tokenanking
Feature activation+0.000
the
Token the
Feature activation+0.000
sc
Token sc
Feature activation+0.000
iss
Tokeniss
Feature activation+0.000
ile
Tokenile
Feature activation+0.000
(
Token (
Feature activation+0.000
Michael
TokenMichael
Feature activation+0.000
C
Token C
Feature activation+0.000
aine
Tokenaine
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
ile
Tokenile
Feature activation+0.000
(
Token (
Feature activation+0.000
Michael
TokenMichael
Feature activation+0.000
C
Token C
Feature activation+0.000
aine
Tokenaine
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
(
Token (
Feature activation+0.000
Michael
TokenMichael
Feature activation+0.000
C
Token C
Feature activation+0.000
aine
Tokenaine
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Michael
TokenMichael
Feature activation+0.000
C
Token C
Feature activation+0.000
aine
Tokenaine
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
anking
Tokenanking
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
anking
Tokenanking
Feature activation+0.000
the
Token the
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
C
Token C
Feature activation+0.000
aine
Tokenaine
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
aine
Tokenaine
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000

Top feature 7 in H0.11: (feature 16453

TOP ACTIVATIONS
MAX = 2.099

<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
like
Token like
Feature activation+0.000
Sh
Token Sh
Feature activation+1.490
ola
Tokenola
Feature activation+0.000
aur
Token aur
Feature activation+0.000
Sh
Token Sh
Feature activation+0.618
ab
Tokenab
Feature activation+0.000
nam
Tokennam
Feature activation+0.000
and
Token and
Feature activation+0.000
A
Token A
Feature activation+0.000
ank
Tokenank
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
ée
Tokenée
Feature activation+0.000
de
Token de
Feature activation+1.202
la
Token la
Feature activation+0.000
Mode
Token Mode
Feature activation+0.000
de
Token de
Feature activation+0.592
la
Token la
Feature activation+0.000
V
Token V
Feature activation+0.000
ille
Tokenille
Feature activation+0.000
de
Token de
Feature activation+0.055
Paris
Token Paris
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
co
Token co
Feature activation+0.548
-
Token-
Feature activation+0.000
re
Tokenre
Feature activation+0.000
former
Tokenformer
Feature activation+0.000
Dr
Token Dr
Feature activation+0.571
.
Token.
Feature activation+0.000
Z
Token Z
Feature activation+0.000
uh
Tokenuh
Feature activation+0.000
di
Tokendi
Feature activation+0.000
J
Token J
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
prof
Token prof
Feature activation+0.146
ect
Tokenect
Feature activation+0.000
o
Tokeno
Feature activation+0.000
,
Token,
Feature activation+0.000
n
Token n
Feature activation+0.450
am
Tokenam
Feature activation+0.000
i
Token i
Feature activation+0.000
am
Tokenam
Feature activation+0.000
pr
Token pr
Feature activation+0.000
idem
Tokenidem
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
in
Token in
Feature activation+0.000
Computer
Token Computer
Feature activation+0.000
world
Tokenworld
Feature activation+0.000
's
Token's
Feature activation+0.000
Sp
Token Sp
Feature activation+0.442
am
Tokenam
Feature activation+0.000
,
Token,
Feature activation+0.000
Mal
Token Mal
Feature activation+0.000
ware
Tokenware
Feature activation+0.000
and
Token and
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
S
Token S
Feature activation+1.411
ull
Tokenull
Feature activation+0.000
est
Tokenest
Feature activation+0.000
's
Token's
Feature activation+0.000
H
Token H
Feature activation+0.387
ars
Tokenars
Feature activation+0.000
Ter
Token Ter
Feature activation+0.000
rain
Tokenrain
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
area
Token area
Feature activation+0.000
were
Token were
Feature activation+0.000
from
Token from
Feature activation+0.000
the
Token the
Feature activation+0.000
N
Token N
Feature activation+0.345
oka
Tokenoka
Feature activation+0.000
totem
Token totem
Feature activation+0.000
(
Token (
Feature activation+0.000
or
Tokenor
Feature activation+0.000
clan
Token clan
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
a
Token a
Feature activation+0.000
controlled
Token controlled
Feature activation+0.000
experiment
Token experiment
Feature activation+0.000
.
Token.
Feature activation+0.000
Sp
Token Sp
Feature activation+0.289
ont
Tokenont
Feature activation+0.000
aneous
Tokenaneous
Feature activation+0.000
generation
Token generation
Feature activation+0.000
,
Token,
Feature activation+0.000
the
Token the
Feature activation+0.000
bus
Token bus
Feature activation+0.000
that
Token that
Feature activation+0.000
never
Token never
Feature activation+0.000
showed
Token showed
Feature activation+0.000
.
Token.
Feature activation+0.000
St
Token St
Feature activation+0.240
.
Token.
Feature activation+0.000
Clair
Token Clair
Feature activation+0.000
scrambled
Token scrambled
Feature activation+0.000
to
Token to
Feature activation+0.000
find
Token find
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
The
TokenThe
Feature activation+0.000
main
Token main
Feature activation+0.000
problem
Token problem
Feature activation+0.000
with
Token with
Feature activation+0.000
Mr
Token Mr
Feature activation+0.231
.
Token.
Feature activation+0.000
Graham
Token Graham
Feature activation+0.000
's
Token's
Feature activation+0.000
approach
Token approach
Feature activation+0.000
is
Token is
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
and
Token and
Feature activation+0.000
allied
Token allied
Feature activation+0.000
groups
Token groups
Feature activation+0.000
like
Token like
Feature activation+0.000
Al
Token Al
Feature activation+0.200
Qaeda
Token Qaeda
Feature activation+0.000
have
Token have
Feature activation+0.000
been
Token been
Feature activation+0.000
based
Token based
Feature activation+0.000
in
Token in
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
between
Token between
Feature activation+0.000
the
Token the
Feature activation+0.000
council
Token council
Feature activation+0.000
and
Token and
Feature activation+0.000
St
Token St
Feature activation+0.197
rim
Tokenrim
Feature activation+0.000
ling
Tokenling
Feature activation+0.000
,
Token,
Feature activation+0.000
who
Token who
Feature activation+0.000
said
Token said
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
role
Token role
Feature activation+0.000
in
Token in
Feature activation+0.000
"
Token "
Feature activation+0.000
Rock
TokenRock
Feature activation+0.000
J
Token J
Feature activation+0.151
ocks
Tokenocks
Feature activation+0.000
".
Token".
Feature activation+0.000
In
Token In
Feature activation+0.000
2012
Token 2012
Feature activation+0.000
,
Token,
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
Publisher
Token Publisher
Feature activation+0.000
of
Token of
Feature activation+0.000
0
Token 0
Feature activation+0.000
to
Token to
Feature activation+0.000
N
Token N
Feature activation+0.150
data
Token data
Feature activation+0.000
signals
Token signals
Feature activation+0.000
with
Token with
Feature activation+0.000
a
Token a
Feature activation+0.000
l
Token l
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
M
Token M
Feature activation+1.771
240
Token240
Feature activation+0.000
B
TokenB
Feature activation+0.000
and
Token and
Feature activation+0.000
M
Token M
Feature activation+0.145
240
Token240
Feature activation+0.000
L
TokenL
Feature activation+0.000
machine
Token machine
Feature activation+0.000
guns
Token guns
Feature activation+0.000
,
Token,
Feature activation+0.000
-
Token-
Feature activation+0.000
modern
Tokenmodern
Feature activation+0.000
international
Token international
Feature activation+0.000
design
Token design
Feature activation+0.000
group
Token group
Feature activation+0.000
Y
Token Y
Feature activation+0.144
oo
Tokenoo
Feature activation+0.000
and
Token and
Feature activation+0.000
Star
Token Star
Feature activation+0.000
ck
Tokenck
Feature activation+0.000
,
Token,
Feature activation+0.000
Jo
Token Jo
Feature activation+0.000
anna
Tokenanna
Feature activation+0.000
Barnes
Token Barnes
Feature activation+0.000
and
Token and
Feature activation+0.000
Jill
Token Jill
Feature activation+0.000
St
Token St
Feature activation+0.141
.
Token.
Feature activation+0.000
John
Token John
Feature activation+0.000
in
Token in
Feature activation+0.000
Hollywood
Token Hollywood
Feature activation+0.000
,
Token,
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
ests
Tokenests
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Changed
TokenChanged
Feature activation+0.000
Sc
Token Sc
Feature activation+0.138
hematic
Tokenhematic
Feature activation+0.000
Vari
Token Vari
Feature activation+0.000
ations
Tokenations
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
G
TokenG
Feature activation+0.172
ed
Tokened
Feature activation+0.000
im
Tokenim
Feature activation+0.000
inas
Tokeninas
Feature activation+0.000
J
Token J
Feature activation+0.138
urg
Tokenurg
Feature activation+0.000
ait
Tokenait
Feature activation+0.000
is
Tokenis
Feature activation+0.000
-
Token -
Feature activation+0.000
bass
Token bass
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
s
Tokens
Feature activation+0.000
10
Token 10
Feature activation+0.000
-
Token-
Feature activation+0.000
speed
Tokenspeed
Feature activation+0.000
De
Token De
Feature activation+0.138
ore
Tokenore
Feature activation+0.000
group
Token group
Feature activation+0.000
and
Token and
Feature activation+0.000
SR
Token SR
Feature activation+0.000
AM
TokenAM
Feature activation+0.000

Top DFA by src position
MAX = 2.640

<|endoftext|>
Token<|endoftext|>
Feature activation+0.241
Top resid features:
like
Token like
Feature activation-0.041
Top resid features:
Sh
Token Sh
Feature activation+1.133
Top resid features:
ola
Tokenola
Feature activation-0.012
Top resid features:
aur
Token aur
Feature activation-0.330
Top resid features:
Sh
Token Sh
Feature activation+1.340
Top resid features:
ab
Tokenab
Feature activation+0.000
Top resid features:
nam
Tokennam
Feature activation+0.000
Top resid features:
and
Token and
Feature activation+0.000
Top resid features:
A
Token A
Feature activation+0.000
Top resid features:
ank
Tokenank
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.451
Top resid features:
ée
Tokenée
Feature activation-0.283
Top resid features:
de
Token de
Feature activation+1.296
Top resid features:
la
Token la
Feature activation-0.173
Top resid features:
Mode
Token Mode
Feature activation-0.047
Top resid features:
de
Token de
Feature activation+1.060
Top resid features:
la
Token la
Feature activation+0.000
Top resid features:
V
Token V
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.268
Top resid features:
co
Token co
Feature activation-0.149
Top resid features:
-
Token-
Feature activation-0.151
Top resid features:
re
Tokenre
Feature activation+0.001
Top resid features:
former
Tokenformer
Feature activation-0.119
Top resid features:
Dr
Token Dr
Feature activation+2.432
Top resid features:
.
Token.
Feature activation+0.000
Top resid features:
Z
Token Z
Feature activation+0.000
Top resid features:
uh
Tokenuh
Feature activation+0.000
Top resid features:
di
Tokendi
Feature activation+0.000
Top resid features:
J
Token J
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.196
Top resid features:
prof
Token prof
Feature activation-0.058
Top resid features:
ect
Tokenect
Feature activation-0.083
Top resid features:
o
Tokeno
Feature activation-0.122
Top resid features:
,
Token,
Feature activation-0.311
Top resid features:
n
Token n
Feature activation+2.541
Top resid features:
am
Tokenam
Feature activation+0.000
Top resid features:
i
Token i
Feature activation+0.000
Top resid features:
am
Tokenam
Feature activation+0.000
Top resid features:
pr
Token pr
Feature activation+0.000
Top resid features:
idem
Tokenidem
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.265
Top resid features:
in
Token in
Feature activation-0.249
Top resid features:
Computer
Token Computer
Feature activation-0.151
Top resid features:
world
Tokenworld
Feature activation-0.212
Top resid features:
's
Token's
Feature activation-0.098
Top resid features:
Sp
Token Sp
Feature activation+2.599
Top resid features:
am
Tokenam
Feature activation+0.000
Top resid features:
,
Token,
Feature activation+0.000
Top resid features:
Mal
Token Mal
Feature activation+0.000
Top resid features:
ware
Tokenware
Feature activation+0.000
Top resid features:
and
Token and
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.311
Top resid features:
S
Token S
Feature activation+0.031
Top resid features:
ull
Tokenull
Feature activation-0.094
Top resid features:
est
Tokenest
Feature activation-0.162
Top resid features:
's
Token's
Feature activation-0.054
Top resid features:
H
Token H
Feature activation+2.067
Top resid features:
ars
Tokenars
Feature activation+0.000
Top resid features:
Ter
Token Ter
Feature activation+0.000
Top resid features:
rain
Tokenrain
Feature activation+0.000
Top resid features:
Ċ
TokenĊ
Feature activation+0.000
Top resid features:
Ċ
TokenĊ
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.249
Top resid features:
area
Token area
Feature activation-0.053
Top resid features:
were
Token were
Feature activation-0.136
Top resid features:
from
Token from
Feature activation-0.157
Top resid features:
the
Token the
Feature activation-0.395
Top resid features:
N
Token N
Feature activation+2.549
Top resid features:
oka
Tokenoka
Feature activation+0.000
Top resid features:
totem
Token totem
Feature activation+0.000
Top resid features:
(
Token (
Feature activation+0.000
Top resid features:
or
Tokenor
Feature activation+0.000
Top resid features:
clan
Token clan
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.271
Top resid features:
a
Token a
Feature activation-0.235
Top resid features:
controlled
Token controlled
Feature activation-0.237
Top resid features:
experiment
Token experiment
Feature activation-0.164
Top resid features:
.
Token.
Feature activation-0.227
Top resid features:
Sp
Token Sp
Feature activation+2.593
Top resid features:
ont
Tokenont
Feature activation+0.000
Top resid features:
aneous
Tokenaneous
Feature activation+0.000
Top resid features:
generation
Token generation
Feature activation+0.000
Top resid features:
,
Token,
Feature activation+0.000
Top resid features:
the
Token the
Feature activation+0.000
Top resid features:
bus
Token bus
Feature activation-0.116
Top resid features:
that
Token that
Feature activation-0.219
Top resid features:
never
Token never
Feature activation-0.169
Top resid features:
showed
Token showed
Feature activation-0.104
Top resid features:
.
Token.
Feature activation-0.203
Top resid features:
St
Token St
Feature activation+2.602
Top resid features:
.
Token.
Feature activation+0.000
Top resid features:
Clair
Token Clair
Feature activation+0.000
Top resid features:
scrambled
Token scrambled
Feature activation+0.000
Top resid features:
to
Token to
Feature activation+0.000
Top resid features:
find
Token find
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.391
Top resid features:
The
TokenThe
Feature activation-0.245
Top resid features:
main
Token main
Feature activation-0.091
Top resid features:
problem
Token problem
Feature activation-0.067
Top resid features:
with
Token with
Feature activation-0.124
Top resid features:
Mr
Token Mr
Feature activation+2.079
Top resid features:
.
Token.
Feature activation+0.000
Top resid features:
Graham
Token Graham
Feature activation+0.000
Top resid features:
's
Token's
Feature activation+0.000
Top resid features:
approach
Token approach
Feature activation+0.000
Top resid features:
is
Token is
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.212
Top resid features:
and
Token and
Feature activation-0.252
Top resid features:
allied
Token allied
Feature activation-0.140
Top resid features:
groups
Token groups
Feature activation-0.131
Top resid features:
like
Token like
Feature activation-0.169
Top resid features:
Al
Token Al
Feature activation+2.391
Top resid features:
Qaeda
Token Qaeda
Feature activation+0.000
Top resid features:
have
Token have
Feature activation+0.000
Top resid features:
been
Token been
Feature activation+0.000
Top resid features:
based
Token based
Feature activation+0.000
Top resid features:
in
Token in
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.240
Top resid features:
between
Token between
Feature activation-0.205
Top resid features:
the
Token the
Feature activation-0.394
Top resid features:
council
Token council
Feature activation-0.098
Top resid features:
and
Token and
Feature activation-0.272
Top resid features:
St
Token St
Feature activation+2.638
Top resid features:
rim
Tokenrim
Feature activation+0.000
Top resid features:
ling
Tokenling
Feature activation+0.000
Top resid features:
,
Token,
Feature activation+0.000
Top resid features:
who
Token who
Feature activation+0.000
Top resid features:
said
Token said
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.325
Top resid features:
role
Token role
Feature activation-0.107
Top resid features:
in
Token in
Feature activation-0.173
Top resid features:
"
Token "
Feature activation-0.059
Top resid features:
Rock
TokenRock
Feature activation-0.213
Top resid features:
J
Token J
Feature activation+2.090
Top resid features:
ocks
Tokenocks
Feature activation+0.000
Top resid features:
".
Token".
Feature activation+0.000
Top resid features:
In
Token In
Feature activation+0.000
Top resid features:
2012
Token 2012
Feature activation+0.000
Top resid features:
,
Token,
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.221
Top resid features:
Publisher
Token Publisher
Feature activation-0.396
Top resid features:
of
Token of
Feature activation-0.159
Top resid features:
0
Token 0
Feature activation-0.122
Top resid features:
to
Token to
Feature activation-0.262
Top resid features:
N
Token N
Feature activation+2.580
Top resid features:
data
Token data
Feature activation+0.000
Top resid features:
signals
Token signals
Feature activation+0.000
Top resid features:
with
Token with
Feature activation+0.000
Top resid features:
a
Token a
Feature activation+0.000
Top resid features:
l
Token l
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.341
Top resid features:
M
Token M
Feature activation+0.859
Top resid features:
240
Token240
Feature activation-0.124
Top resid features:
B
TokenB
Feature activation-0.082
Top resid features:
and
Token and
Feature activation-0.269
Top resid features:
M
Token M
Feature activation+1.131
Top resid features:
240
Token240
Feature activation+0.000
Top resid features:
L
TokenL
Feature activation+0.000
Top resid features:
machine
Token machine
Feature activation+0.000
Top resid features:
guns
Token guns
Feature activation+0.000
Top resid features:
,
Token,
Feature activation+0.000
Top resid features:
-
Token-
Feature activation-0.026
Top resid features:
modern
Tokenmodern
Feature activation-0.126
Top resid features:
international
Token international
Feature activation-0.202
Top resid features:
design
Token design
Feature activation-0.026
Top resid features:
group
Token group
Feature activation-0.036
Top resid features:
Y
Token Y
Feature activation+2.126
Top resid features:
oo
Tokenoo
Feature activation+0.000
Top resid features:
and
Token and
Feature activation+0.000
Top resid features:
Star
Token Star
Feature activation+0.000
Top resid features:
ck
Tokenck
Feature activation+0.000
Top resid features:
,
Token,
Feature activation+0.000
Top resid features:
Jo
Token Jo
Feature activation-0.182
Top resid features:
anna
Tokenanna
Feature activation-0.118
Top resid features:
Barnes
Token Barnes
Feature activation-0.084
Top resid features:
and
Token and
Feature activation-0.283
Top resid features:
Jill
Token Jill
Feature activation-0.245
Top resid features:
St
Token St
Feature activation+2.640
Top resid features:
.
Token.
Feature activation+0.000
Top resid features:
John
Token John
Feature activation+0.000
Top resid features:
in
Token in
Feature activation+0.000
Top resid features:
Hollywood
Token Hollywood
Feature activation+0.000
Top resid features:
,
Token,
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.165
Top resid features:
ests
Tokenests
Feature activation-0.311
Top resid features:
Ċ
TokenĊ
Feature activation-0.181
Top resid features:
Ċ
TokenĊ
Feature activation-0.158
Top resid features:
Changed
TokenChanged
Feature activation-0.248
Top resid features:
Sc
Token Sc
Feature activation+2.585
Top resid features:
hematic
Tokenhematic
Feature activation+0.000
Top resid features:
Vari
Token Vari
Feature activation+0.000
Top resid features:
ations
Tokenations
Feature activation+0.000
Top resid features:
Ċ
TokenĊ
Feature activation+0.000
Top resid features:
Ċ
TokenĊ
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.306
Top resid features:
G
TokenG
Feature activation+0.014
Top resid features:
ed
Tokened
Feature activation-0.171
Top resid features:
im
Tokenim
Feature activation-0.174
Top resid features:
inas
Tokeninas
Feature activation-0.165
Top resid features:
J
Token J
Feature activation+2.040
Top resid features:
urg
Tokenurg
Feature activation+0.000
Top resid features:
ait
Tokenait
Feature activation+0.000
Top resid features:
is
Tokenis
Feature activation+0.000
Top resid features:
-
Token -
Feature activation+0.000
Top resid features:
bass
Token bass
Feature activation+0.000
Top resid features:
Ļ
TokenĻ
Feature activation-0.089
Top resid features:
s
Tokens
Feature activation-0.052
Top resid features:
10
Token 10
Feature activation-0.097
Top resid features:
-
Token-
Feature activation-0.216
Top resid features:
speed
Tokenspeed
Feature activation-0.122
Top resid features:
De
Token De
Feature activation+2.373
Top resid features:
ore
Tokenore
Feature activation+0.000
Top resid features:
group
Token group
Feature activation+0.000
Top resid features:
and
Token and
Feature activation+0.000
Top resid features:
SR
Token SR
Feature activation+0.000
Top resid features:
AM
TokenAM
Feature activation+0.000
Top resid features:

Decoder Weights Distribution

Head 0: 0.06

Head 1: 0.10

Head 2: 0.06

Head 3: 0.08

Head 4: 0.05

Head 5: 0.07

Head 6: 0.11

Head 7: 0.09

Head 8: 0.05

Head 9: 0.14

Head 10: 0.07

Head 11: 0.14

Positive logits

odes3.21

oreal3.21

nutrit3.19

unden3.13

conduc3.10

confir3.04

proble2.98

earchers2.94

mathemat2.92

availability2.90

andem2.88

everal2.85

NetMessage2.84

VIDIA2.83

orry2.78

iner2.75

zbollah2.71

eatures2.68

orem2.68

unte2.61

Negative logits

FactoryReloaded-2.64

edIn-2.50

ーテ-2.43

�士-2.24

��-2.11

}.-2.05

DragonMagazine-2.01

thereafter-1.99

arsity-1.95

BaseType-1.94

afterward-1.94

CLSID-1.91

Tokens-1.84

Mellon-1.84

afterwards-1.83

Treaty-1.83

[_-1.83

Jeb-1.82

shapeshifter-1.82

Havana-1.81

INTERVAL 1.889 - 2.099
CONTAINS 0.001%

INTERVAL 1.680 - 1.889
CONTAINS 0.002%

INTERVAL 1.470 - 1.680
CONTAINS 0.003%

INTERVAL 1.260 - 1.470
CONTAINS 0.003%

INTERVAL 1.050 - 1.260
CONTAINS 0.004%

INTERVAL 0.840 - 1.050
CONTAINS 0.010%

INTERVAL 0.630 - 0.840
CONTAINS 0.014%

INTERVAL 0.420 - 0.630
CONTAINS 0.021%

<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
like
Token like
Feature activation+0.000
Sh
Token Sh
Feature activation+1.490
ola
Tokenola
Feature activation+0.000
aur
Token aur
Feature activation+0.000
Sh
Token Sh
Feature activation+0.618
ab
Tokenab
Feature activation+0.000
nam
Tokennam
Feature activation+0.000
and
Token and
Feature activation+0.000
A
Token A
Feature activation+0.000
ank
Tokenank
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
in
Token in
Feature activation+0.000
Computer
Token Computer
Feature activation+0.000
world
Tokenworld
Feature activation+0.000
's
Token's
Feature activation+0.000
Sp
Token Sp
Feature activation+0.442
am
Tokenam
Feature activation+0.000
,
Token,
Feature activation+0.000
Mal
Token Mal
Feature activation+0.000
ware
Tokenware
Feature activation+0.000
and
Token and
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
ée
Tokenée
Feature activation+0.000
de
Token de
Feature activation+1.202
la
Token la
Feature activation+0.000
Mode
Token Mode
Feature activation+0.000
de
Token de
Feature activation+0.592
la
Token la
Feature activation+0.000
V
Token V
Feature activation+0.000
ille
Tokenille
Feature activation+0.000
de
Token de
Feature activation+0.055
Paris
Token Paris
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
prof
Token prof
Feature activation+0.146
ect
Tokenect
Feature activation+0.000
o
Tokeno
Feature activation+0.000
,
Token,
Feature activation+0.000
n
Token n
Feature activation+0.450
am
Tokenam
Feature activation+0.000
i
Token i
Feature activation+0.000
am
Tokenam
Feature activation+0.000
pr
Token pr
Feature activation+0.000
idem
Tokenidem
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
co
Token co
Feature activation+0.548
-
Token-
Feature activation+0.000
re
Tokenre
Feature activation+0.000
former
Tokenformer
Feature activation+0.000
Dr
Token Dr
Feature activation+0.571
.
Token.
Feature activation+0.000
Z
Token Z
Feature activation+0.000
uh
Tokenuh
Feature activation+0.000
di
Tokendi
Feature activation+0.000
J
Token J
Feature activation+0.000

INTERVAL 0.210 - 0.420
CONTAINS 0.026%

bus
Token bus
Feature activation+0.000
that
Token that
Feature activation+0.000
never
Token never
Feature activation+0.000
showed
Token showed
Feature activation+0.000
.
Token.
Feature activation+0.000
St
Token St
Feature activation+0.240
.
Token.
Feature activation+0.000
Clair
Token Clair
Feature activation+0.000
scrambled
Token scrambled
Feature activation+0.000
to
Token to
Feature activation+0.000
find
Token find
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
a
Token a
Feature activation+0.000
controlled
Token controlled
Feature activation+0.000
experiment
Token experiment
Feature activation+0.000
.
Token.
Feature activation+0.000
Sp
Token Sp
Feature activation+0.289
ont
Tokenont
Feature activation+0.000
aneous
Tokenaneous
Feature activation+0.000
generation
Token generation
Feature activation+0.000
,
Token,
Feature activation+0.000
the
Token the
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
The
TokenThe
Feature activation+0.000
main
Token main
Feature activation+0.000
problem
Token problem
Feature activation+0.000
with
Token with
Feature activation+0.000
Mr
Token Mr
Feature activation+0.231
.
Token.
Feature activation+0.000
Graham
Token Graham
Feature activation+0.000
's
Token's
Feature activation+0.000
approach
Token approach
Feature activation+0.000
is
Token is
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
area
Token area
Feature activation+0.000
were
Token were
Feature activation+0.000
from
Token from
Feature activation+0.000
the
Token the
Feature activation+0.000
N
Token N
Feature activation+0.345
oka
Tokenoka
Feature activation+0.000
totem
Token totem
Feature activation+0.000
(
Token (
Feature activation+0.000
or
Tokenor
Feature activation+0.000
clan
Token clan
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
S
Token S
Feature activation+1.411
ull
Tokenull
Feature activation+0.000
est
Tokenest
Feature activation+0.000
's
Token's
Feature activation+0.000
H
Token H
Feature activation+0.387
ars
Tokenars
Feature activation+0.000
Ter
Token Ter
Feature activation+0.000
rain
Tokenrain
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000

INTERVAL 0.000 - 0.210
CONTAINS 99.916%

personal
Token personal
Feature activation+0.000
profiles
Token profiles
Feature activation+0.000
.
Token.
Feature activation+0.000
The
Token The
Feature activation+0.000
descriptions
Token descriptions
Feature activation+0.000
automatically
Token automatically
Feature activation+0.000
transferred
Token transferred
Feature activation+0.000
to
Token to
Feature activation+0.000
its
Token its
Feature activation+0.000
ad
Token ad
Feature activation+0.000
platform
Token platform
Feature activation+0.000
De
Token De
Feature activation+0.000
ul
Tokenul
Feature activation+0.000
of
Tokenof
Feature activation+0.000
eu
Tokeneu
Feature activation+0.000
and
Token and
Feature activation+0.000
addition
Token addition
Feature activation+0.000
of
Token of
Feature activation+0.000
Bol
Token Bol
Feature activation+0.000
as
Tokenas
Feature activation+0.000
ie
Tokenie
Feature activation+0.000
has
Token has
Feature activation+0.000
the
Token the
Feature activation+0.000
game
Token game
Feature activation+0.000
after
Token after
Feature activation+0.000
the
Token the
Feature activation+0.000
company
Token company
Feature activation+0.000
settled
Token settled
Feature activation+0.000
its
Token its
Feature activation+0.000
dispute
Token dispute
Feature activation+0.000
with
Token with
Feature activation+0.000
Fox
Token Fox
Feature activation+0.000
.
Token.
Feature activation+0.000
the
Token the
Feature activation+0.000
Islamic
Token Islamic
Feature activation+0.000
State
Token State
Feature activation+0.000
,
Token,
Feature activation+0.000
and
Token and
Feature activation+0.000
Clinton
Token Clinton
Feature activation+0.000
âĢ
Token âĢ
Feature activation+0.000
ľ
Tokenľ
Feature activation+0.000
has
Tokenhas
Feature activation+0.000
gone
Token gone
Feature activation+0.000
on
Token on
Feature activation+0.000
iness
Tokeniness
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ŀ
TokenĿ
Feature activation+0.000
you
Token you
Feature activation+0.000
may
Token may
Feature activation+0.000
experience
Token experience
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
sed
Token sed
Feature activation+0.000
uction
Tokenuction
Feature activation+0.000
community
Token community
Feature activation+0.000

BOTTOM ACTIVATIONS
MIN = 0.000

regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
anking
Tokenanking
Feature activation+0.000
the
Token the
Feature activation+0.000
sc
Token sc
Feature activation+0.000
iss
Tokeniss
Feature activation+0.000
ile
Tokenile
Feature activation+0.000
bonds
Token bonds
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
anking
Tokenanking
Feature activation+0.000
the
Token the
Feature activation+0.000
sc
Token sc
Feature activation+0.000
iss
Tokeniss
Feature activation+0.000
ile
Tokenile
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
anking
Tokenanking
Feature activation+0.000
the
Token the
Feature activation+0.000
sc
Token sc
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
anking
Tokenanking
Feature activation+0.000
the
Token the
Feature activation+0.000
sc
Token sc
Feature activation+0.000
iss
Tokeniss
Feature activation+0.000
ile
Tokenile
Feature activation+0.000
(
Token (
Feature activation+0.000
Michael
TokenMichael
Feature activation+0.000
C
Token C
Feature activation+0.000
aine
Tokenaine
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
ile
Tokenile
Feature activation+0.000
(
Token (
Feature activation+0.000
Michael
TokenMichael
Feature activation+0.000
C
Token C
Feature activation+0.000
aine
Tokenaine
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
(
Token (
Feature activation+0.000
Michael
TokenMichael
Feature activation+0.000
C
Token C
Feature activation+0.000
aine
Tokenaine
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Michael
TokenMichael
Feature activation+0.000
C
Token C
Feature activation+0.000
aine
Tokenaine
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
anking
Tokenanking
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
anking
Tokenanking
Feature activation+0.000
the
Token the
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
C
Token C
Feature activation+0.000
aine
Tokenaine
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
aine
Tokenaine
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000

Top feature 8 in H0.11: (feature 20701

TOP ACTIVATIONS
MAX = 1.568

<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
s
Tokens
Feature activation+0.000
the
Token the
Feature activation+0.000
same
Token same
Feature activation+1.010
thing
Token thing
Feature activation+0.000
that
Token that
Feature activation+0.000
worked
Token worked
Feature activation+0.000
to
Token to
Feature activation+0.000
his
Token his
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
ľ
Tokenľ
Feature activation+0.000
Fil
TokenFil
Feature activation+0.000
tering
Tokentering
Feature activation+0.000
certain
Token certain
Feature activation+0.999
content
Token content
Feature activation+0.000
on
Token on
Feature activation+0.000
the
Token the
Feature activation+0.000
web
Token web
Feature activation+0.000
is
Token is
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
coming
Token coming
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
next
Token next
Feature activation+0.000
few
Token few
Feature activation+0.952
weeks
Token weeks
Feature activation+0.000
.
Token.
Feature activation+0.000
If
Token If
Feature activation+0.000
you
Token you
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
,
Token,
Feature activation+0.000
but
Token but
Feature activation+0.000
at
Token at
Feature activation+0.000
the
Token the
Feature activation+0.000
same
Token same
Feature activation+0.924
time
Token time
Feature activation+0.000
it
Token it
Feature activation+0.000
's
Token's
Feature activation+0.000
like
Token like
Feature activation+0.000
a
Token a
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
right
Token right
Feature activation+0.000
out
Token out
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
same
Token same
Feature activation+0.922
adherence
Token adherence
Feature activation+0.000
to
Token to
Feature activation+0.000
19
Token 19
Feature activation+0.000
th
Tokenth
Feature activation+0.000
century
Token century
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
upset
Token upset
Feature activation+0.000
.
Token.
Feature activation+0.000
But
Token But
Feature activation+0.000
very
Token very
Feature activation+0.000
few
Token few
Feature activation+0.896
actually
Token actually
Feature activation+0.000
make
Token make
Feature activation+0.000
me
Token me
Feature activation+0.000
sad
Token sad
Feature activation+0.000
.
Token.
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
that
Token that
Feature activation+0.000
they
Token they
Feature activation+0.000
are
Token are
Feature activation+0.000
exact
Token exact
Feature activation+0.293
same
Token same
Feature activation+0.882
,
Token,
Feature activation+0.000
it
Token it
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
s
Tokens
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
in
Token in
Feature activation+0.000
place
Token place
Feature activation+0.000
for
Token for
Feature activation+0.000
a
Token a
Feature activation+0.000
few
Token few
Feature activation+0.868
years
Token years
Feature activation+0.000
already
Token already
Feature activation+0.000
,
Token,
Feature activation+0.000
since
Token since
Feature activation+0.000
oh
Token oh
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
types
Token types
Feature activation+0.403
of
Token of
Feature activation+0.000
thinking
Token thinking
Feature activation+0.000
and
Token and
Feature activation+0.000
different
Token different
Feature activation+0.862
types
Token types
Feature activation+0.056
of
Token of
Feature activation+0.000
action
Token action
Feature activation+0.000
and
Token and
Feature activation+0.000
response
Token response
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
have
Token have
Feature activation+0.000
location
Token location
Feature activation+0.000
histories
Token histories
Feature activation+0.000
a
Token a
Feature activation+0.000
few
Token few
Feature activation+0.855
hundred
Token hundred
Feature activation+0.000
meg
Token meg
Feature activation+0.000
abytes
Tokenabytes
Feature activation+0.000
in
Token in
Feature activation+0.000
size
Token size
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
programs
Token programs
Feature activation+0.000
,
Token,
Feature activation+0.000
he
Token he
Feature activation+0.000
watches
Token watches
Feature activation+0.000
different
Token different
Feature activation+0.837
interviews
Token interviews
Feature activation+0.000
,
Token,
Feature activation+0.000
he
Token he
Feature activation+0.000
goes
Token goes
Feature activation+0.000
to
Token to
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
There
TokenThere
Feature activation+0.000
are
Token are
Feature activation+0.000
a
Token a
Feature activation+0.000
few
Token few
Feature activation+0.828
downs
Token downs
Feature activation+0.000
ides
Tokenides
Feature activation+0.000
,
Token,
Feature activation+0.000
though
Token though
Feature activation+0.000
.
Token.
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
it
Token it
Feature activation+0.000
le
Token le
Feature activation+0.000
eches
Tokeneches
Feature activation+0.000
a
Token a
Feature activation+0.000
lot
Token lot
Feature activation+0.824
,
Token,
Feature activation+0.000
but
Token but
Feature activation+0.000
only
Token only
Feature activation+0.000
sometimes
Token sometimes
Feature activation+0.000
:
Token:
Feature activation+0.000
Q
TokenQ
Feature activation+0.000
ais
Tokenais
Feature activation+0.000
is
Token is
Feature activation+0.000
not
Token not
Feature activation+0.000
the
Token the
Feature activation+0.000
same
Token same
Feature activation+0.817
as
Token as
Feature activation+0.000
that
Token that
Feature activation+0.000
of
Token of
Feature activation+0.000
Egyptian
Token Egyptian
Feature activation+0.000
poet
Token poet
Feature activation+0.000
aging
Tokenaging
Feature activation+0.000
them
Token them
Feature activation+0.000
)
Token)
Feature activation+0.000
is
Token is
Feature activation+0.000
the
Token the
Feature activation+0.000
same
Token same
Feature activation+0.814
as
Token as
Feature activation+0.000
doing
Token doing
Feature activation+0.000
well
Token well
Feature activation+0.000
(
Token (
Feature activation+0.000
making
Tokenmaking
Feature activation+0.000
come
Token come
Feature activation+0.000
together
Token together
Feature activation+0.000
with
Token with
Feature activation+0.000
the
Token the
Feature activation+0.000
next
Token next
Feature activation+0.000
few
Token few
Feature activation+0.811
months
Token months
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
What
TokenWhat
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
between
Token between
Feature activation+0.000
these
Token these
Feature activation+0.000
two
Token two
Feature activation+0.000
very
Token very
Feature activation+0.210
different
Token different
Feature activation+0.804
disciplines
Token disciplines
Feature activation+0.000
is
Token is
Feature activation+0.000
possible
Token possible
Feature activation+0.000
,
Token,
Feature activation+0.000
one
Token one
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
rock
Tokenrock
Feature activation+0.000
,
Token,
Feature activation+0.000
when
Token when
Feature activation+0.000
a
Token a
Feature activation+0.000
couple
Token couple
Feature activation+0.751
pulled
Token pulled
Feature activation+0.000
over
Token over
Feature activation+0.000
to
Token to
Feature activation+0.000
give
Token give
Feature activation+0.000
him
Token him
Feature activation+0.000
be
Token be
Feature activation+0.000
purchased
Token purchased
Feature activation+0.000
for
Token for
Feature activation+0.000
around
Token around
Feature activation+0.000
the
Token the
Feature activation+0.000
same
Token same
Feature activation+0.736
price
Token price
Feature activation+0.000
as
Token as
Feature activation+0.000
the
Token the
Feature activation+0.000
mould
Token mould
Feature activation+0.000
ing
Tokening
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
to
Token to
Feature activation+0.000
grow
Token grow
Feature activation+0.000
it
Token it
Feature activation+0.000
through
Token through
Feature activation+0.000
various
Token various
Feature activation+0.730
sectors
Token sectors
Feature activation+0.000
such
Token such
Feature activation+0.000
medical
Token medical
Feature activation+0.000
devices
Token devices
Feature activation+0.000
,
Token,
Feature activation+0.000

Top DFA by src position
MAX = 2.008

<|endoftext|>
Token<|endoftext|>
Feature activation+0.456
Top resid features:
âĢ
TokenâĢ
Feature activation+0.022
Top resid features:
Ļ
TokenĻ
Feature activation+0.102
Top resid features:
s
Tokens
Feature activation+0.040
Top resid features:
the
Token the
Feature activation-0.190
Top resid features:
same
Token same
Feature activation+1.956
Top resid features:
thing
Token thing
Feature activation+0.000
Top resid features:
that
Token that
Feature activation+0.000
Top resid features:
worked
Token worked
Feature activation+0.000
Top resid features:
to
Token to
Feature activation+0.000
Top resid features:
his
Token his
Feature activation+0.000
Top resid features:
Ċ
TokenĊ
Feature activation-0.099
Top resid features:
âĢ
TokenâĢ
Feature activation-0.051
Top resid features:
ľ
Tokenľ
Feature activation+0.204
Top resid features:
Fil
TokenFil
Feature activation-0.036
Top resid features:
tering
Tokentering
Feature activation+0.010
Top resid features:
certain
Token certain
Feature activation+1.888
Top resid features:
content
Token content
Feature activation+0.000
Top resid features:
on
Token on
Feature activation+0.000
Top resid features:
the
Token the
Feature activation+0.000
Top resid features:
web
Token web
Feature activation+0.000
Top resid features:
is
Token is
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.415
Top resid features:
coming
Token coming
Feature activation+0.044
Top resid features:
in
Token in
Feature activation-0.087
Top resid features:
the
Token the
Feature activation-0.107
Top resid features:
next
Token next
Feature activation+0.053
Top resid features:
few
Token few
Feature activation+2.008
Top resid features:
weeks
Token weeks
Feature activation+0.000
Top resid features:
.
Token.
Feature activation+0.000
Top resid features:
If
Token If
Feature activation+0.000
Top resid features:
you
Token you
Feature activation+0.000
Top resid features:
âĢ
TokenâĢ
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.576
Top resid features:
,
Token,
Feature activation-0.080
Top resid features:
but
Token but
Feature activation-0.025
Top resid features:
at
Token at
Feature activation-0.017
Top resid features:
the
Token the
Feature activation-0.141
Top resid features:
same
Token same
Feature activation+1.985
Top resid features:
time
Token time
Feature activation+0.000
Top resid features:
it
Token it
Feature activation+0.000
Top resid features:
's
Token's
Feature activation+0.000
Top resid features:
like
Token like
Feature activation+0.000
Top resid features:
a
Token a
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.545
Top resid features:
right
Token right
Feature activation-0.031
Top resid features:
out
Token out
Feature activation+0.034
Top resid features:
of
Token of
Feature activation-0.017
Top resid features:
the
Token the
Feature activation-0.149
Top resid features:
same
Token same
Feature activation+1.915
Top resid features:
adherence
Token adherence
Feature activation+0.000
Top resid features:
to
Token to
Feature activation+0.000
Top resid features:
19
Token 19
Feature activation+0.000
Top resid features:
th
Tokenth
Feature activation+0.000
Top resid features:
century
Token century
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.417
Top resid features:
upset
Token upset
Feature activation+0.111
Top resid features:
.
Token.
Feature activation-0.164
Top resid features:
But
Token But
Feature activation+0.013
Top resid features:
very
Token very
Feature activation+0.040
Top resid features:
few
Token few
Feature activation+1.855
Top resid features:
actually
Token actually
Feature activation+0.000
Top resid features:
make
Token make
Feature activation+0.000
Top resid features:
me
Token me
Feature activation+0.000
Top resid features:
sad
Token sad
Feature activation+0.000
Top resid features:
.
Token.
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.487
Top resid features:
that
Token that
Feature activation-0.111
Top resid features:
they
Token they
Feature activation+0.039
Top resid features:
are
Token are
Feature activation-0.055
Top resid features:
exact
Token exact
Feature activation+0.010
Top resid features:
same
Token same
Feature activation+1.888
Top resid features:
,
Token,
Feature activation+0.000
Top resid features:
it
Token it
Feature activation+0.000
Top resid features:
âĢ
TokenâĢ
Feature activation+0.000
Top resid features:
Ļ
TokenĻ
Feature activation+0.000
Top resid features:
s
Tokens
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.484
Top resid features:
in
Token in
Feature activation-0.102
Top resid features:
place
Token place
Feature activation+0.065
Top resid features:
for
Token for
Feature activation-0.101
Top resid features:
a
Token a
Feature activation-0.099
Top resid features:
few
Token few
Feature activation+1.995
Top resid features:
years
Token years
Feature activation+0.000
Top resid features:
already
Token already
Feature activation+0.000
Top resid features:
,
Token,
Feature activation+0.000
Top resid features:
since
Token since
Feature activation+0.000
Top resid features:
oh
Token oh
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.454
Top resid features:
types
Token types
Feature activation-0.017
Top resid features:
of
Token of
Feature activation-0.064
Top resid features:
thinking
Token thinking
Feature activation+0.120
Top resid features:
and
Token and
Feature activation-0.089
Top resid features:
different
Token different
Feature activation+1.833
Top resid features:
types
Token types
Feature activation+0.000
Top resid features:
of
Token of
Feature activation+0.000
Top resid features:
action
Token action
Feature activation+0.000
Top resid features:
and
Token and
Feature activation+0.000
Top resid features:
response
Token response
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.426
Top resid features:
have
Token have
Feature activation-0.056
Top resid features:
location
Token location
Feature activation+0.061
Top resid features:
histories
Token histories
Feature activation-0.041
Top resid features:
a
Token a
Feature activation-0.121
Top resid features:
few
Token few
Feature activation+1.961
Top resid features:
hundred
Token hundred
Feature activation+0.000
Top resid features:
meg
Token meg
Feature activation+0.000
Top resid features:
abytes
Tokenabytes
Feature activation+0.000
Top resid features:
in
Token in
Feature activation+0.000
Top resid features:
size
Token size
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.461
Top resid features:
programs
Token programs
Feature activation+0.071
Top resid features:
,
Token,
Feature activation-0.093
Top resid features:
he
Token he
Feature activation-0.096
Top resid features:
watches
Token watches
Feature activation+0.013
Top resid features:
different
Token different
Feature activation+1.857
Top resid features:
interviews
Token interviews
Feature activation+0.000
Top resid features:
,
Token,
Feature activation+0.000
Top resid features:
he
Token he
Feature activation+0.000
Top resid features:
goes
Token goes
Feature activation+0.000
Top resid features:
to
Token to
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.532
Top resid features:
Ċ
TokenĊ
Feature activation-0.111
Top resid features:
There
TokenThere
Feature activation-0.045
Top resid features:
are
Token are
Feature activation-0.080
Top resid features:
a
Token a
Feature activation-0.101
Top resid features:
few
Token few
Feature activation+2.008
Top resid features:
downs
Token downs
Feature activation+0.000
Top resid features:
ides
Tokenides
Feature activation+0.000
Top resid features:
,
Token,
Feature activation+0.000
Top resid features:
though
Token though
Feature activation+0.000
Top resid features:
.
Token.
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.527
Top resid features:
it
Token it
Feature activation+0.036
Top resid features:
le
Token le
Feature activation-0.042
Top resid features:
eches
Tokeneches
Feature activation+0.024
Top resid features:
a
Token a
Feature activation-0.082
Top resid features:
lot
Token lot
Feature activation+1.736
Top resid features:
,
Token,
Feature activation+0.000
Top resid features:
but
Token but
Feature activation+0.000
Top resid features:
only
Token only
Feature activation+0.000
Top resid features:
sometimes
Token sometimes
Feature activation+0.000
Top resid features:
:
Token:
Feature activation+0.000
Top resid features:
Q
TokenQ
Feature activation+0.057
Top resid features:
ais
Tokenais
Feature activation+0.034
Top resid features:
is
Token is
Feature activation-0.086
Top resid features:
not
Token not
Feature activation-0.006
Top resid features:
the
Token the
Feature activation-0.164
Top resid features:
same
Token same
Feature activation+1.874
Top resid features:
as
Token as
Feature activation+0.000
Top resid features:
that
Token that
Feature activation+0.000
Top resid features:
of
Token of
Feature activation+0.000
Top resid features:
Egyptian
Token Egyptian
Feature activation+0.000
Top resid features:
poet
Token poet
Feature activation+0.000
Top resid features:
aging
Tokenaging
Feature activation+0.068
Top resid features:
them
Token them
Feature activation+0.075
Top resid features:
)
Token)
Feature activation-0.077
Top resid features:
is
Token is
Feature activation-0.073
Top resid features:
the
Token the
Feature activation-0.149
Top resid features:
same
Token same
Feature activation+1.894
Top resid features:
as
Token as
Feature activation+0.000
Top resid features:
doing
Token doing
Feature activation+0.000
Top resid features:
well
Token well
Feature activation+0.000
Top resid features:
(
Token (
Feature activation+0.000
Top resid features:
making
Tokenmaking
Feature activation+0.000
Top resid features:
come
Token come
Feature activation+0.010
Top resid features:
together
Token together
Feature activation-0.041
Top resid features:
with
Token with
Feature activation-0.107
Top resid features:
the
Token the
Feature activation-0.106
Top resid features:
next
Token next
Feature activation+0.052
Top resid features:
few
Token few
Feature activation+1.959
Top resid features:
months
Token months
Feature activation+0.000
Top resid features:
.
Token.
Feature activation+0.000
Top resid features:
Ċ
TokenĊ
Feature activation+0.000
Top resid features:
Ċ
TokenĊ
Feature activation+0.000
Top resid features:
What
TokenWhat
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.577
Top resid features:
between
Token between
Feature activation-0.094
Top resid features:
these
Token these
Feature activation+0.041
Top resid features:
two
Token two
Feature activation-0.038
Top resid features:
very
Token very
Feature activation-0.079
Top resid features:
different
Token different
Feature activation+1.771
Top resid features:
disciplines
Token disciplines
Feature activation+0.000
Top resid features:
is
Token is
Feature activation+0.000
Top resid features:
possible
Token possible
Feature activation+0.000
Top resid features:
,
Token,
Feature activation+0.000
Top resid features:
one
Token one
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.489
Top resid features:
rock
Tokenrock
Feature activation+0.073
Top resid features:
,
Token,
Feature activation-0.079
Top resid features:
when
Token when
Feature activation-0.050
Top resid features:
a
Token a
Feature activation-0.059
Top resid features:
couple
Token couple
Feature activation+1.752
Top resid features:
pulled
Token pulled
Feature activation+0.000
Top resid features:
over
Token over
Feature activation+0.000
Top resid features:
to
Token to
Feature activation+0.000
Top resid features:
give
Token give
Feature activation+0.000
Top resid features:
him
Token him
Feature activation+0.000
Top resid features:
be
Token be
Feature activation-0.022
Top resid features:
purchased
Token purchased
Feature activation+0.062
Top resid features:
for
Token for
Feature activation-0.055
Top resid features:
around
Token around
Feature activation-0.025
Top resid features:
the
Token the
Feature activation-0.161
Top resid features:
same
Token same
Feature activation+1.893
Top resid features:
price
Token price
Feature activation+0.000
Top resid features:
as
Token as
Feature activation+0.000
Top resid features:
the
Token the
Feature activation+0.000
Top resid features:
mould
Token mould
Feature activation+0.000
Top resid features:
ing
Tokening
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.509
Top resid features:
to
Token to
Feature activation-0.075
Top resid features:
grow
Token grow
Feature activation+0.062
Top resid features:
it
Token it
Feature activation-0.020
Top resid features:
through
Token through
Feature activation-0.064
Top resid features:
various
Token various
Feature activation+1.692
Top resid features:
sectors
Token sectors
Feature activation+0.000
Top resid features:
such
Token such
Feature activation+0.000
Top resid features:
medical
Token medical
Feature activation+0.000
Top resid features:
devices
Token devices
Feature activation+0.000
Top resid features:
,
Token,
Feature activation+0.000
Top resid features:

Decoder Weights Distribution

Head 0: 0.06

Head 1: 0.10

Head 2: 0.07

Head 3: 0.10

Head 4: 0.04

Head 5: 0.06

Head 6: 0.09

Head 7: 0.07

Head 8: 0.06

Head 9: 0.15

Head 10: 0.07

Head 11: 0.14

Positive logits

proble3.43

NetMessage3.16

lihood2.97

terday2.80

NESS2.77

etheless2.75

vre2.73

unden2.73

ibilities2.68

orry2.67

endiary2.61

confir2.56

orem2.54

2.53

aukee2.52

hess2.47

nutrit2.46

otine2.45

conduc2.43

SourceFile2.39

Negative logits

FactoryReloaded-2.29

illions-2.17

Tokens-2.01

arsity-1.98

Laksh-1.93

town-1.90

[|-1.81

Tracker-1.73

Discussion-1.73

CLSID-1.72

built-1.70

alli-1.69

}.-1.68

afterward-1.68

459-1.67

alf-1.67

them-1.65

borrowed-1.65

960-1.64

Gang-1.64

INTERVAL 1.411 - 1.568
CONTAINS 0.002%

INTERVAL 1.254 - 1.411
CONTAINS 0.001%

INTERVAL 1.097 - 1.254
CONTAINS 0.004%

INTERVAL 0.941 - 1.097
CONTAINS 0.005%

<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
s
Tokens
Feature activation+0.000
the
Token the
Feature activation+0.000
same
Token same
Feature activation+1.010
thing
Token thing
Feature activation+0.000
that
Token that
Feature activation+0.000
worked
Token worked
Feature activation+0.000
to
Token to
Feature activation+0.000
his
Token his
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
ľ
Tokenľ
Feature activation+0.000
Fil
TokenFil
Feature activation+0.000
tering
Tokentering
Feature activation+0.000
certain
Token certain
Feature activation+0.999
content
Token content
Feature activation+0.000
on
Token on
Feature activation+0.000
the
Token the
Feature activation+0.000
web
Token web
Feature activation+0.000
is
Token is
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
coming
Token coming
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
next
Token next
Feature activation+0.000
few
Token few
Feature activation+0.952
weeks
Token weeks
Feature activation+0.000
.
Token.
Feature activation+0.000
If
Token If
Feature activation+0.000
you
Token you
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000

INTERVAL 0.784 - 0.941
CONTAINS 0.011%

<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
,
Token,
Feature activation+0.000
but
Token but
Feature activation+0.000
at
Token at
Feature activation+0.000
the
Token the
Feature activation+0.000
same
Token same
Feature activation+0.924
time
Token time
Feature activation+0.000
it
Token it
Feature activation+0.000
's
Token's
Feature activation+0.000
like
Token like
Feature activation+0.000
a
Token a
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
between
Token between
Feature activation+0.000
these
Token these
Feature activation+0.000
two
Token two
Feature activation+0.000
very
Token very
Feature activation+0.210
different
Token different
Feature activation+0.804
disciplines
Token disciplines
Feature activation+0.000
is
Token is
Feature activation+0.000
possible
Token possible
Feature activation+0.000
,
Token,
Feature activation+0.000
one
Token one
Feature activation+0.000
Q
TokenQ
Feature activation+0.000
ais
Tokenais
Feature activation+0.000
is
Token is
Feature activation+0.000
not
Token not
Feature activation+0.000
the
Token the
Feature activation+0.000
same
Token same
Feature activation+0.817
as
Token as
Feature activation+0.000
that
Token that
Feature activation+0.000
of
Token of
Feature activation+0.000
Egyptian
Token Egyptian
Feature activation+0.000
poet
Token poet
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
programs
Token programs
Feature activation+0.000
,
Token,
Feature activation+0.000
he
Token he
Feature activation+0.000
watches
Token watches
Feature activation+0.000
different
Token different
Feature activation+0.837
interviews
Token interviews
Feature activation+0.000
,
Token,
Feature activation+0.000
he
Token he
Feature activation+0.000
goes
Token goes
Feature activation+0.000
to
Token to
Feature activation+0.000
aging
Tokenaging
Feature activation+0.000
them
Token them
Feature activation+0.000
)
Token)
Feature activation+0.000
is
Token is
Feature activation+0.000
the
Token the
Feature activation+0.000
same
Token same
Feature activation+0.814
as
Token as
Feature activation+0.000
doing
Token doing
Feature activation+0.000
well
Token well
Feature activation+0.000
(
Token (
Feature activation+0.000
making
Tokenmaking
Feature activation+0.000

INTERVAL 0.627 - 0.784
CONTAINS 0.012%

<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
,
Token,
Feature activation+0.000
and
Token and
Feature activation+0.000
a
Token a
Feature activation+0.000
potential
Token potential
Feature activation+0.284
full
Token full
Feature activation+0.668
closure
Token closure
Feature activation+0.000
will
Token will
Feature activation+0.000
come
Token come
Feature activation+0.000
as
Token as
Feature activation+0.000
a
Token a
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
to
Token to
Feature activation+0.000
grow
Token grow
Feature activation+0.000
it
Token it
Feature activation+0.000
through
Token through
Feature activation+0.000
various
Token various
Feature activation+0.730
sectors
Token sectors
Feature activation+0.000
such
Token such
Feature activation+0.000
medical
Token medical
Feature activation+0.000
devices
Token devices
Feature activation+0.000
,
Token,
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
rock
Tokenrock
Feature activation+0.000
,
Token,
Feature activation+0.000
when
Token when
Feature activation+0.000
a
Token a
Feature activation+0.000
couple
Token couple
Feature activation+0.751
pulled
Token pulled
Feature activation+0.000
over
Token over
Feature activation+0.000
to
Token to
Feature activation+0.000
give
Token give
Feature activation+0.000
him
Token him
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
a
Token a
Feature activation+0.000
result
Token result
Feature activation+0.111
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
lot
Token lot
Feature activation+0.694
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
messaging
Token messaging
Feature activation+0.000
we
Token we
Feature activation+0.000
used
Token used
Feature activation+0.000
staying
Token staying
Feature activation+0.000
in
Token in
Feature activation+0.000
touch
Token touch
Feature activation+0.000
with
Token with
Feature activation+0.000
a
Token a
Feature activation+0.000
few
Token few
Feature activation+0.692
friends
Token friends
Feature activation+0.000
who
Token who
Feature activation+0.000
have
Token have
Feature activation+0.000
similar
Token similar
Feature activation+0.000
interests
Token interests
Feature activation+0.000

INTERVAL 0.470 - 0.627
CONTAINS 0.018%

¦
Token¦
Feature activation+0.344
á
Tokená
Feature activation+0.000
ĵ
Tokenĵ
Feature activation+0.980
¯
Token¯
Feature activation+0.000
á
Tokená
Feature activation+0.000
ĵ
Tokenĵ
Feature activation+0.627
Ĥ
TokenĤ
Feature activation+0.000
á
Tokená
Feature activation+0.000
ķ
Tokenķ
Feature activation+0.000
Ī
TokenĪ
Feature activation+0.000
á
Tokená
Feature activation+0.000
s
Tokens
Feature activation+0.000
comes
Token comes
Feature activation+0.000
out
Token out
Feature activation+0.000
of
Token of
Feature activation+0.000
a
Token a
Feature activation+0.000
different
Token different
Feature activation+0.605
world
Token world
Feature activation+0.000
,
Token,
Feature activation+0.000
and
Token and
Feature activation+0.000
getting
Token getting
Feature activation+0.000
back
Token back
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
ll
Tokenll
Feature activation+0.000
be
Token be
Feature activation+0.000
good
Token good
Feature activation+0.500
.
Token.
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ŀ
TokenĿ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
Corps
Token Corps
Feature activation+0.000
Dr
Token Dr
Feature activation+0.000
.,
Token.,
Feature activation+0.000
the
Token the
Feature activation+0.000
main
Token main
Feature activation+0.514
drag
Token drag
Feature activation+0.000
.
Token.
Feature activation+0.000
"
Token "
Feature activation+0.000
Not
TokenNot
Feature activation+0.000
beautiful
Token beautiful
Feature activation+0.000
it
Token it
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
s
Tokens
Feature activation+0.000
not
Token not
Feature activation+0.000
true
Token true
Feature activation+0.526
.
Token.
Feature activation+0.000
According
Token According
Feature activation+0.000
to
Token to
Feature activation+0.000
the
Token the
Feature activation+0.000
company
Token company
Feature activation+0.000

INTERVAL 0.314 - 0.470
CONTAINS 0.020%

<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
viruses
Token viruses
Feature activation+0.000
to
Token to
Feature activation+0.000
destroy
Token destroy
Feature activation+0.000
various
Token various
Feature activation+0.815
types
Token types
Feature activation+0.335
of
Token of
Feature activation+0.000
cancer
Token cancer
Feature activation+0.000
over
Token over
Feature activation+0.000
the
Token the
Feature activation+0.000
years
Token years
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
We
TokenWe
Feature activation+0.000
've
Token've
Feature activation+0.000
missed
Token missed
Feature activation+0.000
the
Token the
Feature activation+0.000
entire
Token entire
Feature activation+0.381
purpose
Token purpose
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
End
Token End
Feature activation+0.000
angered
Tokenangered
Feature activation+0.000
wife
Token wife
Feature activation+0.000
is
Token is
Feature activation+0.000
a
Token a
Feature activation+0.000
pre
Token pre
Feature activation+0.000
law
Tokenlaw
Feature activation+0.000
major
Token major
Feature activation+0.359
at
Token at
Feature activation+0.000
UN
Token UN
Feature activation+0.000
LV
TokenLV
Feature activation+0.000
.
Token.
Feature activation+0.000
Pretty
Token Pretty
Feature activation+0.000
make
Token make
Feature activation+0.000
it
Token it
Feature activation+0.000
in
Token in
Feature activation+0.000
his
Token his
Feature activation+0.000
last
Token last
Feature activation+0.000
couple
Token couple
Feature activation+0.340
of
Token of
Feature activation+0.000
years
Token years
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
and
Token and
Feature activation+0.000
thought
Token thought
Feature activation+0.000
captured
Token captured
Feature activation+0.000
in
Token in
Feature activation+0.000
these
Token these
Feature activation+0.000
few
Token few
Feature activation+0.388
essays
Token essays
Feature activation+0.000
.
Token.
Feature activation+0.000
What
Token What
Feature activation+0.000
Brian
Token Brian
Feature activation+0.000
Kern
Token Kern
Feature activation+0.000

INTERVAL 0.157 - 0.314
CONTAINS 0.029%

<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
tempo
Token tempo
Feature activation+0.000
,
Token,
Feature activation+0.000
per
Token per
Feature activation+0.000
Ã
TokenÃ
Feature activation+0.000
²
Token²
Feature activation+0.278
,
Token,
Feature activation+0.000
le
Token le
Feature activation+0.000
inform
Token inform
Feature activation+0.000
az
Tokenaz
Feature activation+0.000
ion
Tokenion
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
This
TokenThis
Feature activation+0.000
book
Token book
Feature activation+0.008
has
Token has
Feature activation+0.000
a
Token a
Feature activation+0.000
great
Token great
Feature activation+0.271
pace
Token pace
Feature activation+0.000
and
Token and
Feature activation+0.000
I
Token I
Feature activation+0.000
was
Token was
Feature activation+0.000
hooked
Token hooked
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
-
Token-
Feature activation+0.000
Mal
TokenMal
Feature activation+0.000
iki
Tokeniki
Feature activation+0.000
a
Token a
Feature activation+0.000
great
Token great
Feature activation+0.223
classical
Token classical
Feature activation+0.000
scholar
Token scholar
Feature activation+0.000
,
Token,
Feature activation+0.000
wrote
Token wrote
Feature activation+0.000
,
Token,
Feature activation+0.000
.
Token.
Feature activation+0.000
m
Tokenm
Feature activation+0.000
.
Token.
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
As
TokenAs
Feature activation+0.000
good
Token good
Feature activation+0.300
as
Token as
Feature activation+0.000
Google
Token Google
Feature activation+0.000
Maps
Token Maps
Feature activation+0.000
is
Token is
Feature activation+0.000
for
Token for
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
These
TokenThese
Feature activation+0.000
are
Token are
Feature activation+0.000
really
Token really
Feature activation+0.000
good
Token good
Feature activation+0.281
.
Token.
Feature activation+0.000
I
Token I
Feature activation+0.000
experimented
Token experimented
Feature activation+0.000
with
Token with
Feature activation+0.000
my
Token my
Feature activation+0.000

INTERVAL 0.000 - 0.157
CONTAINS 99.898%

22
Token 22
Feature activation+0.000
members
Token members
Feature activation+0.000
out
Token out
Feature activation+0.000
of
Token of
Feature activation+0.000
1300
Token 1300
Feature activation+0.000
members
Token members
Feature activation+0.000
to
Token to
Feature activation+0.000
be
Token be
Feature activation+0.000
nationally
Token nationally
Feature activation+0.000
certified
Token certified
Feature activation+0.000
by
Token by
Feature activation+0.000
riders
Token riders
Feature activation+0.000
commonly
Token commonly
Feature activation+0.000
have
Token have
Feature activation+0.000
few
Token few
Feature activation+0.000
(
Token (
Feature activation+0.000
or
Tokenor
Feature activation+0.000
no
Token no
Feature activation+0.000
)
Token)
Feature activation+0.000
good
Token good
Feature activation+0.000
options
Token options
Feature activation+0.000
for
Token for
Feature activation+0.000
associated
Token associated
Feature activation+0.000
with
Token with
Feature activation+0.000
war
Token war
Feature activation+0.000
lords
Tokenlords
Feature activation+0.000
committing
Token committing
Feature activation+0.000
atrocities
Token atrocities
Feature activation+0.000
-
Token -
Feature activation+0.000
and
Token and
Feature activation+0.000
more
Token more
Feature activation+0.000
to
Token to
Feature activation+0.000
the
Token the
Feature activation+0.000
make
Token make
Feature activation+0.000
news
Token news
Feature activation+0.000
consumption
Token consumption
Feature activation+0.000
on
Token on
Feature activation+0.000
Facebook
Token Facebook
Feature activation+0.000
even
Token even
Feature activation+0.000
easier
Token easier
Feature activation+0.000
,
Token,
Feature activation+0.000
including
Token including
Feature activation+0.000
Instant
Token Instant
Feature activation+0.000
Articles
Token Articles
Feature activation+0.000
a
Token a
Feature activation+0.000
few
Token few
Feature activation+0.000
examples
Token examples
Feature activation+0.000
of
Token of
Feature activation+0.000
emerging
Token emerging
Feature activation+0.000
technologies
Token technologies
Feature activation+0.000
criminals
Token criminals
Feature activation+0.000
could
Token could
Feature activation+0.000
exploit
Token exploit
Feature activation+0.000
for
Token for
Feature activation+0.000
deadly
Token deadly
Feature activation+0.000

BOTTOM ACTIVATIONS
MIN = 0.000

regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
anking
Tokenanking
Feature activation+0.000
the
Token the
Feature activation+0.000
sc
Token sc
Feature activation+0.000
iss
Tokeniss
Feature activation+0.000
ile
Tokenile
Feature activation+0.000
bonds
Token bonds
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
anking
Tokenanking
Feature activation+0.000
the
Token the
Feature activation+0.000
sc
Token sc
Feature activation+0.000
iss
Tokeniss
Feature activation+0.000
ile
Tokenile
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
anking
Tokenanking
Feature activation+0.000
the
Token the
Feature activation+0.000
sc
Token sc
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
anking
Tokenanking
Feature activation+0.000
the
Token the
Feature activation+0.000
sc
Token sc
Feature activation+0.000
iss
Tokeniss
Feature activation+0.000
ile
Tokenile
Feature activation+0.000
(
Token (
Feature activation+0.000
Michael
TokenMichael
Feature activation+0.000
C
Token C
Feature activation+0.000
aine
Tokenaine
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
ile
Tokenile
Feature activation+0.000
(
Token (
Feature activation+0.000
Michael
TokenMichael
Feature activation+0.000
C
Token C
Feature activation+0.000
aine
Tokenaine
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
(
Token (
Feature activation+0.000
Michael
TokenMichael
Feature activation+0.000
C
Token C
Feature activation+0.000
aine
Tokenaine
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Michael
TokenMichael
Feature activation+0.000
C
Token C
Feature activation+0.000
aine
Tokenaine
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
anking
Tokenanking
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
anking
Tokenanking
Feature activation+0.000
the
Token the
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
C
Token C
Feature activation+0.000
aine
Tokenaine
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
aine
Tokenaine
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000

Top feature 9 in H0.11: (feature 8183

TOP ACTIVATIONS
MAX = 2.599

<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
ko
Tokenko
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
s
Tokens
Feature activation+0.000
first
Token first
Feature activation+1.090
appearance
Token appearance
Feature activation+0.000
at
Token at
Feature activation+0.000
the
Token the
Feature activation+0.000
Mart
Token Mart
Feature activation+0.000
ins
Tokenins
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
aren
Tokenaren
Feature activation+0.000
took
Token took
Feature activation+0.000
on
Token on
Feature activation+0.000
its
Token its
Feature activation+0.000
first
Token first
Feature activation+1.025
mobile
Token mobile
Feature activation+0.000
project
Token project
Feature activation+0.000
using
Token using
Feature activation+0.000
React
Token React
Feature activation+0.000
Native
TokenNative
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
this
Token this
Feature activation+0.000
sense
Token sense
Feature activation+0.000
of
Token of
Feature activation+0.000
anarchy
Token anarchy
Feature activation+0.000
first
Token first
Feature activation+0.976
-
Token-
Feature activation+0.000
hand
Tokenhand
Feature activation+0.000
:
Token:
Feature activation+0.000
from
Token from
Feature activation+0.000
1981
Token 1981
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
The
TokenThe
Feature activation+0.000
case
Token case
Feature activation+0.000
is
Token is
Feature activation+0.000
the
Token the
Feature activation+0.000
first
Token first
Feature activation+0.965
to
Token to
Feature activation+0.000
be
Token be
Feature activation+0.000
brought
Token brought
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
ready
Token ready
Feature activation+0.000
to
Token to
Feature activation+0.000
build
Token build
Feature activation+0.000
your
Token your
Feature activation+0.000
first
Token first
Feature activation+0.923
Android
Token Android
Feature activation+0.000
app
Token app
Feature activation+0.000
?
Token?
Feature activation+0.000
Great
Token Great
Feature activation+0.000
choice
Token choice
Feature activation+0.000
first
Token first
Feature activation+2.502
time
Token time
Feature activation+0.000
since
Token since
Feature activation+0.000
their
Token their
Feature activation+0.000
names
Token names
Feature activation+0.000
first
Token first
Feature activation+0.889
appeared
Token appeared
Feature activation+0.000
on
Token on
Feature activation+0.000
the
Token the
Feature activation+0.000
ballot
Token ballot
Feature activation+0.000
in
Token in
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
sell
Token sell
Feature activation+0.000
it
Token it
Feature activation+0.000
.
Token.
Feature activation+0.000
At
Token At
Feature activation+0.000
first
Token first
Feature activation+0.849
glance
Token glance
Feature activation+0.000
,
Token,
Feature activation+0.000
that
Token that
Feature activation+0.000
would
Token would
Feature activation+0.000
make
Token make
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
beaten
Token beaten
Feature activation+0.000
is
Token is
Feature activation+0.000
not
Token not
Feature activation+0.000
the
Token the
Feature activation+0.000
first
Token first
Feature activation+0.768
thing
Token thing
Feature activation+0.000
on
Token on
Feature activation+0.000
anybody
Token anybody
Feature activation+0.000
's
Token's
Feature activation+0.000
mind
Token mind
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
,
Token,
Feature activation+0.000
Oregon
Token Oregon
Feature activation+0.000
became
Token became
Feature activation+0.000
the
Token the
Feature activation+0.000
first
Token first
Feature activation+0.759
state
Token state
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
nation
Token nation
Feature activation+0.000
to
Token to
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
contract
Token contract
Feature activation+0.000
to
Token to
Feature activation+0.000
build
Token build
Feature activation+0.000
the
Token the
Feature activation+0.000
first
Token first
Feature activation+0.751
phase
Token phase
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
Domain
Token Domain
Feature activation+0.000
Awareness
Token Awareness
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
bill
Token bill
Feature activation+0.000
that
Token that
Feature activation+0.000
was
Token was
Feature activation+0.000
first
Token first
Feature activation+0.684
presented
Token presented
Feature activation+0.000
in
Token in
Feature activation+0.000
2001
Token 2001
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
,
Token,
Feature activation+0.000
where
Token where
Feature activation+0.000
it
Token it
Feature activation+0.000
built
Token built
Feature activation+0.000
its
Token its
Feature activation+0.000
first
Token first
Feature activation+0.676
church
Token church
Feature activation+0.000
in
Token in
Feature activation+0.000
N
Token N
Feature activation+0.000
airo
Tokenairo
Feature activation+0.000
bi
Tokenbi
Feature activation+0.000
foster
Token foster
Feature activation+0.000
care
Token care
Feature activation+0.000
r
Tokenr
Feature activation+0.000
,
Token,
Feature activation+0.000
was
Token was
Feature activation+0.000
first
Token first
Feature activation+0.547
contacted
Token contacted
Feature activation+0.000
by
Token by
Feature activation+0.000
police
Token police
Feature activation+0.000
he
Token he
Feature activation+0.000
was
Token was
Feature activation+0.000
years
Token years
Feature activation+0.000
ago
Token ago
Feature activation+0.000
I
Token I
Feature activation+0.000
bought
Token bought
Feature activation+0.000
my
Token my
Feature activation+0.000
first
Token first
Feature activation+0.540
consumer
Token consumer
Feature activation+0.000
knitting
Token knitting
Feature activation+0.000
machine
Token machine
Feature activation+0.000
,
Token,
Feature activation+0.000
modified
Token modified
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
ths
Tokenths
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
First
TokenFirst
Feature activation+0.522
and
Token and
Feature activation+0.000
most
Token most
Feature activation+0.000
importantly
Token importantly
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
t
Tokent
Feature activation+0.000
run
Token run
Feature activation+0.000
alone
Token alone
Feature activation+0.000
after
Token after
Feature activation+0.000
the
Token the
Feature activation+0.000
first
Token first
Feature activation+0.513
fascist
Token fascist
Feature activation+0.000
you
Token you
Feature activation+0.000
see
Token see
Feature activation+0.000
,
Token,
Feature activation+0.000
but
Token but
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
into
Token into
Feature activation+0.000
halves
Token halves
Feature activation+0.000
.
Token.
Feature activation+0.000
The
Token The
Feature activation+0.000
first
Token first
Feature activation+0.493
âĢ
Token âĢ
Feature activation+0.000
ľ
Tokenľ
Feature activation+0.000
half
Tokenhalf
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ŀ
TokenĿ
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
said
Token said
Feature activation+0.000
.
Token.
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
The
TokenThe
Feature activation+0.000
second
Token second
Feature activation+0.476
video
Token video
Feature activation+0.000
in
Token in
Feature activation+0.000
James
Token James
Feature activation+0.000
O
Token O
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
National
Token National
Feature activation+0.000
Football
Token Football
Feature activation+0.000
League
Token League
Feature activation+0.000
advertisement
Token advertisement
Feature activation+0.000
that
Token that
Feature activation+0.000
first
Token first
Feature activation+0.467
aired
Token aired
Feature activation+0.000
during
Token during
Feature activation+0.000
Super
Token Super
Feature activation+0.000
Bowl
Token Bowl
Feature activation+0.000
XL
Token XL
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
the
Token the
Feature activation+0.000
Canucks
Token Canucks
Feature activation+0.000
have
Token have
Feature activation+0.000
the
Token the
Feature activation+0.000
second
Token second
Feature activation+0.447
-
Token-
Feature activation+0.000
worst
Tokenworst
Feature activation+0.000
luck
Token luck
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000

Top DFA by src position
MAX = 6.198

<|endoftext|>
Token<|endoftext|>
Feature activation+1.636
Top resid features:
ko
Tokenko
Feature activation+0.109
Top resid features:
âĢ
TokenâĢ
Feature activation-0.164
Top resid features:
Ļ
TokenĻ
Feature activation+0.061
Top resid features:
s
Tokens
Feature activation+0.315
Top resid features:
first
Token first
Feature activation+6.198
Top resid features:
appearance
Token appearance
Feature activation+0.000
Top resid features:
at
Token at
Feature activation+0.000
Top resid features:
the
Token the
Feature activation+0.000
Top resid features:
Mart
Token Mart
Feature activation+0.000
Top resid features:
ins
Tokenins
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+1.281
Top resid features:
aren
Tokenaren
Feature activation-0.004
Top resid features:
took
Token took
Feature activation+0.182
Top resid features:
on
Token on
Feature activation+0.252
Top resid features:
its
Token its
Feature activation+0.338
Top resid features:
first
Token first
Feature activation+6.041
Top resid features:
mobile
Token mobile
Feature activation+0.000
Top resid features:
project
Token project
Feature activation+0.000
Top resid features:
using
Token using
Feature activation+0.000
Top resid features:
React
Token React
Feature activation+0.000
Top resid features:
Native
TokenNative
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+1.180
Top resid features:
this
Token this
Feature activation+0.498
Top resid features:
sense
Token sense
Feature activation+0.046
Top resid features:
of
Token of
Feature activation+0.325
Top resid features:
anarchy
Token anarchy
Feature activation-0.039
Top resid features:
first
Token first
Feature activation+6.030
Top resid features:
-
Token-
Feature activation+0.000
Top resid features:
hand
Tokenhand
Feature activation+0.000
Top resid features:
:
Token:
Feature activation+0.000
Top resid features:
from
Token from
Feature activation+0.000
Top resid features:
1981
Token 1981
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+1.126
Top resid features:
The
TokenThe
Feature activation+0.332
Top resid features:
case
Token case
Feature activation+0.026
Top resid features:
is
Token is
Feature activation+0.220
Top resid features:
the
Token the
Feature activation+0.374
Top resid features:
first
Token first
Feature activation+5.951
Top resid features:
to
Token to
Feature activation+0.000
Top resid features:
be
Token be
Feature activation+0.000
Top resid features:
brought
Token brought
Feature activation+0.000
Top resid features:
by
Token by
Feature activation+0.000
Top resid features:
the
Token the
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+1.315
Top resid features:
ready
Token ready
Feature activation+0.141
Top resid features:
to
Token to
Feature activation+0.329
Top resid features:
build
Token build
Feature activation-0.129
Top resid features:
your
Token your
Feature activation+0.339
Top resid features:
first
Token first
Feature activation+5.991
Top resid features:
Android
Token Android
Feature activation+0.000
Top resid features:
app
Token app
Feature activation+0.000
Top resid features:
?
Token?
Feature activation+0.000
Top resid features:
Great
Token Great
Feature activation+0.000
Top resid features:
choice
Token choice
Feature activation+0.000
Top resid features:
first
Token first
Feature activation+2.082
Top resid features:
time
Token time
Feature activation+0.140
Top resid features:
since
Token since
Feature activation+0.058
Top resid features:
their
Token their
Feature activation+0.235
Top resid features:
names
Token names
Feature activation-0.038
Top resid features:
first
Token first
Feature activation+4.055
Top resid features:
appeared
Token appeared
Feature activation+0.000
Top resid features:
on
Token on
Feature activation+0.000
Top resid features:
the
Token the
Feature activation+0.000
Top resid features:
ballot
Token ballot
Feature activation+0.000
Top resid features:
in
Token in
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+1.142
Top resid features:
sell
Token sell
Feature activation+0.072
Top resid features:
it
Token it
Feature activation+0.277
Top resid features:
.
Token.
Feature activation+0.134
Top resid features:
At
Token At
Feature activation+0.320
Top resid features:
first
Token first
Feature activation+5.970
Top resid features:
glance
Token glance
Feature activation+0.000
Top resid features:
,
Token,
Feature activation+0.000
Top resid features:
that
Token that
Feature activation+0.000
Top resid features:
would
Token would
Feature activation+0.000
Top resid features:
make
Token make
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+1.108
Top resid features:
beaten
Token beaten
Feature activation-0.065
Top resid features:
is
Token is
Feature activation+0.282
Top resid features:
not
Token not
Feature activation+0.067
Top resid features:
the
Token the
Feature activation+0.417
Top resid features:
first
Token first
Feature activation+6.024
Top resid features:
thing
Token thing
Feature activation+0.000
Top resid features:
on
Token on
Feature activation+0.000
Top resid features:
anybody
Token anybody
Feature activation+0.000
Top resid features:
's
Token's
Feature activation+0.000
Top resid features:
mind
Token mind
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+1.082
Top resid features:
,
Token,
Feature activation+0.347
Top resid features:
Oregon
Token Oregon
Feature activation-0.112
Top resid features:
became
Token became
Feature activation+0.124
Top resid features:
the
Token the
Feature activation+0.434
Top resid features:
first
Token first
Feature activation+5.949
Top resid features:
state
Token state
Feature activation+0.000
Top resid features:
in
Token in
Feature activation+0.000
Top resid features:
the
Token the
Feature activation+0.000
Top resid features:
nation
Token nation
Feature activation+0.000
Top resid features:
to
Token to
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+1.161
Top resid features:
contract
Token contract
Feature activation+0.141
Top resid features:
to
Token to
Feature activation+0.264
Top resid features:
build
Token build
Feature activation-0.149
Top resid features:
the
Token the
Feature activation+0.419
Top resid features:
first
Token first
Feature activation+5.980
Top resid features:
phase
Token phase
Feature activation+0.000
Top resid features:
of
Token of
Feature activation+0.000
Top resid features:
the
Token the
Feature activation+0.000
Top resid features:
Domain
Token Domain
Feature activation+0.000
Top resid features:
Awareness
Token Awareness
Feature activation+0.000
Top resid features:
,
Token,
Feature activation+0.260
Top resid features:
a
Token a
Feature activation+0.389
Top resid features:
bill
Token bill
Feature activation+0.114
Top resid features:
that
Token that
Feature activation+0.093
Top resid features:
was
Token was
Feature activation+0.178
Top resid features:
first
Token first
Feature activation+5.770
Top resid features:
presented
Token presented
Feature activation+0.000
Top resid features:
in
Token in
Feature activation+0.000
Top resid features:
2001
Token 2001
Feature activation+0.000
Top resid features:
.
Token.
Feature activation+0.000
Top resid features:
Ċ
TokenĊ
Feature activation+0.000
Top resid features:
,
Token,
Feature activation+0.316
Top resid features:
where
Token where
Feature activation+0.094
Top resid features:
it
Token it
Feature activation+0.203
Top resid features:
built
Token built
Feature activation-0.012
Top resid features:
its
Token its
Feature activation+0.264
Top resid features:
first
Token first
Feature activation+5.782
Top resid features:
church
Token church
Feature activation+0.000
Top resid features:
in
Token in
Feature activation+0.000
Top resid features:
N
Token N
Feature activation+0.000
Top resid features:
airo
Tokenairo
Feature activation+0.000
Top resid features:
bi
Tokenbi
Feature activation+0.000
Top resid features:
foster
Token foster
Feature activation+0.009
Top resid features:
care
Token care
Feature activation+0.052
Top resid features:
r
Tokenr
Feature activation+0.143
Top resid features:
,
Token,
Feature activation+0.179
Top resid features:
was
Token was
Feature activation+0.251
Top resid features:
first
Token first
Feature activation+5.909
Top resid features:
contacted
Token contacted
Feature activation+0.000
Top resid features:
by
Token by
Feature activation+0.000
Top resid features:
police
Token police
Feature activation+0.000
Top resid features:
he
Token he
Feature activation+0.000
Top resid features:
was
Token was
Feature activation+0.000
Top resid features:
years
Token years
Feature activation+0.276
Top resid features:
ago
Token ago
Feature activation+0.021
Top resid features:
I
Token I
Feature activation+0.249
Top resid features:
bought
Token bought
Feature activation-0.068
Top resid features:
my
Token my
Feature activation+0.275
Top resid features:
first
Token first
Feature activation+5.835
Top resid features:
consumer
Token consumer
Feature activation+0.000
Top resid features:
knitting
Token knitting
Feature activation+0.000
Top resid features:
machine
Token machine
Feature activation+0.000
Top resid features:
,
Token,
Feature activation+0.000
Top resid features:
modified
Token modified
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+1.801
Top resid features:
ths
Tokenths
Feature activation+0.151
Top resid features:
.
Token.
Feature activation+0.228
Top resid features:
Ċ
TokenĊ
Feature activation+0.110
Top resid features:
Ċ
TokenĊ
Feature activation+0.074
Top resid features:
First
TokenFirst
Feature activation+5.223
Top resid features:
and
Token and
Feature activation+0.000
Top resid features:
most
Token most
Feature activation+0.000
Top resid features:
importantly
Token importantly
Feature activation+0.000
Top resid features:
,
Token,
Feature activation+0.000
Top resid features:
a
Token a
Feature activation+0.000
Top resid features:
t
Tokent
Feature activation+0.258
Top resid features:
run
Token run
Feature activation+0.019
Top resid features:
alone
Token alone
Feature activation+0.027
Top resid features:
after
Token after
Feature activation+0.032
Top resid features:
the
Token the
Feature activation+0.399
Top resid features:
first
Token first
Feature activation+5.768
Top resid features:
fascist
Token fascist
Feature activation+0.000
Top resid features:
you
Token you
Feature activation+0.000
Top resid features:
see
Token see
Feature activation+0.000
Top resid features:
,
Token,
Feature activation+0.000
Top resid features:
but
Token but
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+1.159
Top resid features:
into
Token into
Feature activation+0.091
Top resid features:
halves
Token halves
Feature activation-0.185
Top resid features:
.
Token.
Feature activation+0.129
Top resid features:
The
Token The
Feature activation+0.348
Top resid features:
first
Token first
Feature activation+6.015
Top resid features:
âĢ
Token âĢ
Feature activation+0.000
Top resid features:
ľ
Tokenľ
Feature activation+0.000
Top resid features:
half
Tokenhalf
Feature activation+0.000
Top resid features:
âĢ
TokenâĢ
Feature activation+0.000
Top resid features:
Ŀ
TokenĿ
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+1.229
Top resid features:
said
Token said
Feature activation+0.266
Top resid features:
.
Token.
Feature activation+0.189
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+0.245
Top resid features:
The
TokenThe
Feature activation+0.212
Top resid features:
second
Token second
Feature activation+5.400
Top resid features:
video
Token video
Feature activation+0.000
Top resid features:
in
Token in
Feature activation+0.000
Top resid features:
James
Token James
Feature activation+0.000
Top resid features:
O
Token O
Feature activation+0.000
Top resid features:
âĢ
TokenâĢ
Feature activation+0.000
Top resid features:
National
Token National
Feature activation+0.363
Top resid features:
Football
Token Football
Feature activation+0.132
Top resid features:
League
Token League
Feature activation+0.074
Top resid features:
advertisement
Token advertisement
Feature activation-0.083
Top resid features:
that
Token that
Feature activation+0.185
Top resid features:
first
Token first
Feature activation+5.872
Top resid features:
aired
Token aired
Feature activation+0.000
Top resid features:
during
Token during
Feature activation+0.000
Top resid features:
Super
Token Super
Feature activation+0.000
Top resid features:
Bowl
Token Bowl
Feature activation+0.000
Top resid features:
XL
Token XL
Feature activation+0.000
Top resid features:
<|endoftext|>
Token<|endoftext|>
Feature activation+1.192
Top resid features:
the
Token the
Feature activation+0.491
Top resid features:
Canucks
Token Canucks
Feature activation-0.004
Top resid features:
have
Token have
Feature activation+0.064
Top resid features:
the
Token the
Feature activation+0.300
Top resid features:
second
Token second
Feature activation+5.469
Top resid features:
-
Token-
Feature activation+0.000
Top resid features:
worst
Tokenworst
Feature activation+0.000
Top resid features:
luck
Token luck
Feature activation+0.000
Top resid features:
in
Token in
Feature activation+0.000
Top resid features:
the
Token the
Feature activation+0.000
Top resid features:

Decoder Weights Distribution

Head 0: 0.06

Head 1: 0.12

Head 2: 0.06

Head 3: 0.11

Head 4: 0.05

Head 5: 0.06

Head 6: 0.09

Head 7: 0.06

Head 8: 0.05

Head 9: 0.13

Head 10: 0.07

Head 11: 0.13

Positive logits

NetMessage3.83

lihood3.59

etheless3.34

proble2.99

confir2.98

suspic2.84

esson2.81

unden2.77

answ2.69

otine2.67

pmwiki2.65

endiary2.62

aukee2.59

princ2.58

terday2.52

nutshell2.50

conservancy2.48

mite2.47

appre2.45

zbollah2.41

Negative logits

[|-2.39

]}-2.28

}"-2.26

town-2.23

}.-2.20

rooms-2.14

illions-2.13

IQ-2.12

FactoryReloaded-2.07

sym-2.03

eful-1.95

pheus-1.94

Polo-1.94

master-1.93

arsity-1.93

Sut-1.90

chuk-1.88

}\-1.86

-1.86

masters-1.85

INTERVAL 2.339 - 2.599
CONTAINS 0.001%

INTERVAL 2.079 - 2.339
CONTAINS 0.000%

INTERVAL 1.819 - 2.079
CONTAINS 0.001%

INTERVAL 1.559 - 1.819
CONTAINS 0.000%

INTERVAL 1.299 - 1.559
CONTAINS 0.001%

INTERVAL 1.039 - 1.299
CONTAINS 0.001%

<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
ko
Tokenko
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
s
Tokens
Feature activation+0.000
first
Token first
Feature activation+1.090
appearance
Token appearance
Feature activation+0.000
at
Token at
Feature activation+0.000
the
Token the
Feature activation+0.000
Mart
Token Mart
Feature activation+0.000
ins
Tokenins
Feature activation+0.000

INTERVAL 0.780 - 1.039
CONTAINS 0.001%

<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
The
TokenThe
Feature activation+0.000
case
Token case
Feature activation+0.000
is
Token is
Feature activation+0.000
the
Token the
Feature activation+0.000
first
Token first
Feature activation+0.965
to
Token to
Feature activation+0.000
be
Token be
Feature activation+0.000
brought
Token brought
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
ready
Token ready
Feature activation+0.000
to
Token to
Feature activation+0.000
build
Token build
Feature activation+0.000
your
Token your
Feature activation+0.000
first
Token first
Feature activation+0.923
Android
Token Android
Feature activation+0.000
app
Token app
Feature activation+0.000
?
Token?
Feature activation+0.000
Great
Token Great
Feature activation+0.000
choice
Token choice
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
aren
Tokenaren
Feature activation+0.000
took
Token took
Feature activation+0.000
on
Token on
Feature activation+0.000
its
Token its
Feature activation+0.000
first
Token first
Feature activation+1.025
mobile
Token mobile
Feature activation+0.000
project
Token project
Feature activation+0.000
using
Token using
Feature activation+0.000
React
Token React
Feature activation+0.000
Native
TokenNative
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
this
Token this
Feature activation+0.000
sense
Token sense
Feature activation+0.000
of
Token of
Feature activation+0.000
anarchy
Token anarchy
Feature activation+0.000
first
Token first
Feature activation+0.976
-
Token-
Feature activation+0.000
hand
Tokenhand
Feature activation+0.000
:
Token:
Feature activation+0.000
from
Token from
Feature activation+0.000
1981
Token 1981
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
sell
Token sell
Feature activation+0.000
it
Token it
Feature activation+0.000
.
Token.
Feature activation+0.000
At
Token At
Feature activation+0.000
first
Token first
Feature activation+0.849
glance
Token glance
Feature activation+0.000
,
Token,
Feature activation+0.000
that
Token that
Feature activation+0.000
would
Token would
Feature activation+0.000
make
Token make
Feature activation+0.000

INTERVAL 0.520 - 0.780
CONTAINS 0.001%

<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
contract
Token contract
Feature activation+0.000
to
Token to
Feature activation+0.000
build
Token build
Feature activation+0.000
the
Token the
Feature activation+0.000
first
Token first
Feature activation+0.751
phase
Token phase
Feature activation+0.000
of
Token of
Feature activation+0.000
the
Token the
Feature activation+0.000
Domain
Token Domain
Feature activation+0.000
Awareness
Token Awareness
Feature activation+0.000
years
Token years
Feature activation+0.000
ago
Token ago
Feature activation+0.000
I
Token I
Feature activation+0.000
bought
Token bought
Feature activation+0.000
my
Token my
Feature activation+0.000
first
Token first
Feature activation+0.540
consumer
Token consumer
Feature activation+0.000
knitting
Token knitting
Feature activation+0.000
machine
Token machine
Feature activation+0.000
,
Token,
Feature activation+0.000
modified
Token modified
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
beaten
Token beaten
Feature activation+0.000
is
Token is
Feature activation+0.000
not
Token not
Feature activation+0.000
the
Token the
Feature activation+0.000
first
Token first
Feature activation+0.768
thing
Token thing
Feature activation+0.000
on
Token on
Feature activation+0.000
anybody
Token anybody
Feature activation+0.000
's
Token's
Feature activation+0.000
mind
Token mind
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
ths
Tokenths
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
First
TokenFirst
Feature activation+0.522
and
Token and
Feature activation+0.000
most
Token most
Feature activation+0.000
importantly
Token importantly
Feature activation+0.000
,
Token,
Feature activation+0.000
a
Token a
Feature activation+0.000
,
Token,
Feature activation+0.000
where
Token where
Feature activation+0.000
it
Token it
Feature activation+0.000
built
Token built
Feature activation+0.000
its
Token its
Feature activation+0.000
first
Token first
Feature activation+0.676
church
Token church
Feature activation+0.000
in
Token in
Feature activation+0.000
N
Token N
Feature activation+0.000
airo
Tokenairo
Feature activation+0.000
bi
Tokenbi
Feature activation+0.000

INTERVAL 0.260 - 0.520
CONTAINS 0.001%

<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
the
Token the
Feature activation+0.000
Canucks
Token Canucks
Feature activation+0.000
have
Token have
Feature activation+0.000
the
Token the
Feature activation+0.000
second
Token second
Feature activation+0.447
-
Token-
Feature activation+0.000
worst
Tokenworst
Feature activation+0.000
luck
Token luck
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
courtesy
Token courtesy
Feature activation+0.000
Facebook
Token Facebook
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
first
Token first
Feature activation+0.285
question
Token question
Feature activation+0.000
Facebook
Token Facebook
Feature activation+0.000
âĢ
TokenâĢ
Feature activation+0.000
Ļ
TokenĻ
Feature activation+0.000
s
Tokens
Feature activation+0.000
t
Tokent
Feature activation+0.000
run
Token run
Feature activation+0.000
alone
Token alone
Feature activation+0.000
after
Token after
Feature activation+0.000
the
Token the
Feature activation+0.000
first
Token first
Feature activation+0.513
fascist
Token fascist
Feature activation+0.000
you
Token you
Feature activation+0.000
see
Token see
Feature activation+0.000
,
Token,
Feature activation+0.000
but
Token but
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
SEC
Token SEC
Feature activation+0.000
team
Token team
Feature activation+0.000
championships
Token championships
Feature activation+0.000
ranks
Token ranks
Feature activation+0.000
second
Token second
Feature activation+0.336
amongst
Token amongst
Feature activation+0.000
all
Token all
Feature activation+0.000
sports
Token sports
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
National
Token National
Feature activation+0.000
Football
Token Football
Feature activation+0.000
League
Token League
Feature activation+0.000
advertisement
Token advertisement
Feature activation+0.000
that
Token that
Feature activation+0.000
first
Token first
Feature activation+0.467
aired
Token aired
Feature activation+0.000
during
Token during
Feature activation+0.000
Super
Token Super
Feature activation+0.000
Bowl
Token Bowl
Feature activation+0.000
XL
Token XL
Feature activation+0.000

INTERVAL 0.000 - 0.260
CONTAINS 99.993%

.
Token.
Feature activation+0.000
This
Token This
Feature activation+0.000
is
Token is
Feature activation+0.000
not
Token not
Feature activation+0.000
a
Token a
Feature activation+0.000
game
Token game
Feature activation+0.000
that
Token that
Feature activation+0.000
doubles
Token doubles
Feature activation+0.000
down
Token down
Feature activation+0.000
on
Token on
Feature activation+0.000
the
Token the
Feature activation+0.000
a
Token a
Feature activation+0.000
ruling
Token ruling
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
Ontario
Token Ontario
Feature activation+0.000
Superior
Token Superior
Feature activation+0.000
Court
Token Court
Feature activation+0.000
.
Token.
Feature activation+0.000
Justice
Token Justice
Feature activation+0.000
Kim
Token Kim
Feature activation+0.000
Carpenter
Token Carpenter
Feature activation+0.000
is
Token is
Feature activation+0.000
as
Token as
Feature activation+0.000
inflation
Token inflation
Feature activation+0.000
starts
Token starts
Feature activation+0.000
running
Token running
Feature activation+0.000
out
Token out
Feature activation+0.000
of
Token of
Feature activation+0.000
control
Token control
Feature activation+0.000
and
Token and
Feature activation+0.000
prices
Token prices
Feature activation+0.000
start
Token start
Feature activation+0.000
affects
Token affects
Feature activation+0.000
minorities
Token minorities
Feature activation+0.000
.
Token.
Feature activation+0.000
Moreover
Token Moreover
Feature activation+0.000
,
Token,
Feature activation+0.000
Austin
Token Austin
Feature activation+0.000
's
Token's
Feature activation+0.000
history
Token history
Feature activation+0.000
of
Token of
Feature activation+0.000
segregation
Token segregation
Feature activation+0.000
concentrated
Token concentrated
Feature activation+0.000
conviction
Token conviction
Feature activation+0.000
,
Token,
Feature activation+0.000
said
Token said
Feature activation+0.000
prosecutor
Token prosecutor
Feature activation+0.000
David
Token David
Feature activation+0.000
De
Token De
Feature activation+0.000
akin
Tokenakin
Feature activation+0.000
,
Token,
Feature activation+0.000
who
Token who
Feature activation+0.000
called
Token called
Feature activation+0.000
her
Token her
Feature activation+0.000

BOTTOM ACTIVATIONS
MIN = 0.000

regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
anking
Tokenanking
Feature activation+0.000
the
Token the
Feature activation+0.000
sc
Token sc
Feature activation+0.000
iss
Tokeniss
Feature activation+0.000
ile
Tokenile
Feature activation+0.000
bonds
Token bonds
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
anking
Tokenanking
Feature activation+0.000
the
Token the
Feature activation+0.000
sc
Token sc
Feature activation+0.000
iss
Tokeniss
Feature activation+0.000
ile
Tokenile
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
anking
Tokenanking
Feature activation+0.000
the
Token the
Feature activation+0.000
sc
Token sc
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
anking
Tokenanking
Feature activation+0.000
the
Token the
Feature activation+0.000
sc
Token sc
Feature activation+0.000
iss
Tokeniss
Feature activation+0.000
ile
Tokenile
Feature activation+0.000
(
Token (
Feature activation+0.000
Michael
TokenMichael
Feature activation+0.000
C
Token C
Feature activation+0.000
aine
Tokenaine
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
<|endoftext|>
Token<|endoftext|>
Feature activation+0.000
ile
Tokenile
Feature activation+0.000
(
Token (
Feature activation+0.000
Michael
TokenMichael
Feature activation+0.000
C
Token C
Feature activation+0.000
aine
Tokenaine
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
(
Token (
Feature activation+0.000
Michael
TokenMichael
Feature activation+0.000
C
Token C
Feature activation+0.000
aine
Tokenaine
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Michael
TokenMichael
Feature activation+0.000
C
Token C
Feature activation+0.000
aine
Tokenaine
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
anking
Tokenanking
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
is
Token is
Feature activation+0.000
regulated
Token regulated
Feature activation+0.000
by
Token by
Feature activation+0.000
the
Token the
Feature activation+0.000
residues
Token residues
Feature activation+0.000
fl
Token fl
Feature activation+0.000
anking
Tokenanking
Feature activation+0.000
the
Token the
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
ag
Tokenag
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000
G
Token G
Feature activation+0.000
C
Token C
Feature activation+0.000
aine
Tokenaine
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
aine
Tokenaine
Feature activation+0.000
)
Token)
Feature activation+0.000
in
Token in
Feature activation+0.000
the
Token the
Feature activation+0.000
process
Token process
Feature activation+0.000
.
Token.
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
Ċ
TokenĊ
Feature activation+0.000
The
TokenThe
Feature activation+0.000
processing
Token processing
Feature activation+0.000
of
Token of
Feature activation+0.000